6.18
The Biological Pump in the Past
D. M. Sigman
Princeton University, NJ, USA
and
G. H. Haug
Geoforschungszentrum Potsdam, Germany
6.18.1 INTRODUCTION
6.18.2 CONCEPTS
6.18.2.1 Low- and Mid-latitude Ocean
6.18.2.2 High-latitude Ocean
6.18.3 TOOLS
6.18.3.1 Export Production
6.18.3.2 Nutrient Status
6.18.3.3 Integrative Constraints on the Biological Pump
6.18.3.3.1 Carbon isotope distribution of the ocean and atmosphere
6.18.3.3.2 Deep-ocean oxygen content
6.18.3.3.3 Phasing
6.18.4 OBSERVATIONS
6.18.4.1 Low- and Mid-latitude Ocean
6.18.4.2 High-latitude Ocean
6.18.5 SUMMARY AND CURRENT OPINION
ACKNOWLEDGMENTS
REFERENCES
6.18.1 INTRODUCTION
491
496
499
503
509
509
511
512
512
513
513
514
514
517
520
522
522
pump is actually more specific: the sequestration
of carbon dioxide in the ocean interior by the
biogenic flux of organic matter out of surface
waters and into the deep sea prior to decomposition of that organic matter back to carbon dioxide
(Volk and Hoffert, 1985) (Figure 1). The biological pump extracts carbon from the “surface skin”
of the ocean that interacts with the atmosphere,
presenting a lower partial pressure of carbon
dioxide (CO2) to the atmosphere and thus lowering its CO2 content.
The place of the biological pump in the global
carbon cycle is illustrated in Figure 2. The
atmosphere exchanges carbon with essentially
three reservoirs: the ocean, the terrestrial biosphere, and the geosphere. The ocean holds ,50
times as much carbon as does the atmosphere, and
It is easy to imagine that the terrestrial
biosphere sequesters atmospheric carbon dioxide;
the form and quantity of the sequestered carbon,
living or dead organic matter, are striking. In the
ocean, there are no aggregations of biomass
comparable to the forests on land. Yet biological
productivity in the ocean plays a central role in the
sequestration of atmospheric carbon dioxide,
typically overshadowing the effects of terrestrial
biospheric carbon storage on timescales longer
than a few centuries. In an effort to communicate
the ocean’s role in the regulation of atmospheric
carbon dioxide, marine scientists frequently refer
to the ocean’s biologically driven sequestration of
carbon as the “biological pump.” The original and
strict definition of the biological (or “soft-tissue”)
491
492
The Biological Pump in the Past
Figure 1 A schematic of the ocean’s “biological
pump,” the sequestration of carbon and nutrients
(nitrogen and phosphorus) in the ocean interior (lower
dark blue box) by the growth of phytoplankton (floating
unicellular algae) in the sunlit surface ocean (upper light
blue box), the downward rain of organic matter out of
the surface ocean and into the deep ocean (green
downward arrow), and the subsequent breakdown of
this organic matter back to carbon dioxide (CO2), nitrate
32
ðNO2
3 Þ, and phosphate ðPO4 Þ. The nutrients and CO2
are reintroduced to the surface ocean by mixing and
upwelling (the circling arrows at left). The biological
pump lowers the CO2 content of the atmosphere by
extracting it from the surface ocean (which exchanges
CO2 with the atmosphere) and sequestering it in the
isolated waters of the ocean interior. In most of the lowand mid-latitude ocean, the surface is isolated from the
deep sea by a temperature-driven density gradient, or
“thermocline,” keeping nutrient supply low and leading
to essentially complete consumption of NO2
3 and
PO32
4 by phytoplankton at the surface. At the higher
latitudes, where there is no permanent thermocline,
more rapid communication with the deep sea leads to
32
incomplete consumption of NO2
3 and PO4 .
,20 times as much as the terrestrial biosphere.
Thus, on timescales that are adequately long to
allow the deep sea to communicate with the
surface ocean and atmosphere (*500 yr), carbon
storage in the ocean all but sets the concentration
of CO2 in the atmosphere. The effect of the
biological pump is not permanent on this timescale of ocean circulation; the downward transport
of carbon is balanced by the net upward flux from
the CO2-rich waters of the deep sea, which, in the
absence of the biological pump, would work to
homogenize the carbon chemistry of the ocean,
raising atmospheric CO2 in the process.
The dynamics of marine organic carbon can be
described as a set of three nested cycles in which
the biological pump is the cycle with flux and
reservoir of intermediate magnitude (Figure 3).
The cycle with the shortest timescale, which
operates within the surface ocean, is composed
of net primary production by phytoplankton
(their photosynthesis less their respiration) and
heterotrophic respiration by zooplankton and
bacteria that oxidize most of the net primary
production back to CO2. This cycle is by far the
greatest in terms of the flux of carbon, but the
reservoir of sequestered carbon that accumulates
in surface waters (phytoplankton biomass, dead
organic particles, and dissolved organic carbon) is
small relative to the atmospheric reservoir of CO2,
and it has a short residence time (less than a year).
A small imbalance in this cycle, between net
primary production and heterotrophic respiration,
feeds the next cycle in the form of organic carbon
that sinks (or is mixed) into the deep sea. This
“export production” drives the biological pump
(Figure 3).
At the other extreme, a small fraction of the
organic matter exported from the surface ocean
escapes remineralization in the water column and
sediments and is thus buried in the sediments,
removing carbon from the ocean/atmosphere
system (Figure 3). On the timescale of geologic
processes, this carbon removal is balanced by
the oxidation of the organic matter when it is
exposed at the earth surface by uplift and weathering, or when it is released by metamorphism
and volcanism. While the fluxes involved in this
cycle are small, the reservoir is large, so that its
importance increases with timescale, becoming
clearly relevant on the timescale of millions
of years.
The biological pump, in the strict sense, does
not include the burial of organic matter in marine
sediments. There are several related reasons for
this exclusion. First, as described below, the
variations in atmospheric carbon dioxide that are
correlated with the waxing and waning of ice ages
have driven much of the thinking about the role of
organic carbon production on atmospheric composition. Only a very small fraction of the organic
matter exported out of surface waters is preserved
and buried in the underlying sediment, so that this
process cannot sequester a significant amount of
carbon on the timescale of millennia and thus is
not a candidate process for the major carbon
dioxide variations over glacial cycles. Second, on
the timescale for which organic matter burial is
relevant, i.e., over millions of years, it is thought
to be only one of several mechanisms by which
atmospheric CO2 is regulated. Most hypotheses
regarding the history of CO2 over geological time
involve weathering of rocks on land and the
precipitation of carbonates in the ocean, due to
both the larger fluxes and larger reservoirs
involved in the geological trapping of CO2 as
carbonate as opposed to organic carbon (Figure 2).
Finally, the importance of biological productivity
in determining the burial rate of organic carbon is
not at all clear. While some examples of high
organic carbon burial appear to be due to high
Introduction
493
Figure 2 A simplified view of the Holocene (pre-industrial) carbon cycle (largely based on Holmen (1992)). Units
of carbon are petagrams (Pg, 1015 g). The fluxes are colored according to the residence time of carbon in the
reservoirs they involve, from shortest to longest, as follows: red, orange, purple, blue. Exchanges of the atmosphere
with the surface ocean and terrestrial biosphere are relatively rapid, such that changes in the fluxes alter atmospheric
CO2 on the timescale of years and decades. Exchange between the surface ocean and the deep ocean is such that
centuries to millennia are required for a change to yield a new steady state. Because the deep-ocean carbon reservoir
is large relative to the surface ocean, terrestrial biosphere, and atmosphere, its interactions with the surface ocean,
given thousands of years, determine the total amount of carbon to be partitioned among those three reservoirs. In a
lifeless ocean, the only cause for gradients in CO2 between the deep ocean and the surface ocean would be
temperature (i.e., the “solubility pump” (Volk and Hoffert, 1985))—deep waters are formed from especially cold
surface waters, and CO2 is more soluble in cold water. The biological pump, represented by the right downward
arrow, greatly enhances the surface– deep CO2 gradient, through the rain of organic matter (“Corg”) out of the surface
ocean. The “carbonate pump,” represented by the left downward arrow, is the downward rain of calcium carbonate
microfossils out of the surface ocean. Its effect is to actually raise the pCO2 of the surface ocean; this involves the
alkalinity of seawater and is discussed in Chapters 6.04, 6.10, 6.19, and 6.20. Almost all of the organic matter raining
out of the surface ocean is degraded back to CO2 and inorganic nutrients as it rains to the seafloor or once it is
incorporated into the shallow sediments; only less than 1% (,0.05 Pg out of ,10 Pg) is removed from the ocean/
atmosphere system by burial (the downward blue arrow). This is in contrast to the calcium carbonate rain out of the
surface (the downward purple arrow), ,25% of which is buried. In parallel, the weathering rate of calcium carbonate
on land is significant on millennial timescales (the left-pointing purple arrow), while the release of geologically
sequestered Corg occurs more slowly, on a similar timescale as the release of CO2 from the metamorphism of calcium
carbonate to silicate minerals.
oceanic productivity (Vincent and Berger, 1985),
others involve diverse additional processes, such
as the delivery of sediments to the ocean
(FranceLanord and Derry, 1997), which influences the ease with which the organic matter rain
enters the sedimentary record (Canfield, 1994). In
the latter case, the distribution of oceanic productivity is of secondary importance to the effect of
organic carbon burial on atmospheric CO2.
The most direct evidence for natural variations
in atmospheric CO2 comes from the air that is
trapped in Antarctic glacial ice. Records
from Antarctic ice cores indicate that the concentration of CO2 in the atmosphere has varied
in step with the waxing and waning of ice
ages (Barnola et al., 1987; Petit et al., 1999)
(Figure 4). During interglacial times, such as the
Holocene (roughly the past 104 years), the
atmospheric partial pressure of CO2 ( pCO2) is
typically near 280 ppm by volume (ppmv). During
peak glacial times, such as the last glacial
maximum ,1.8 £ 104 yr ago, atmospheric pCO2
494
The Biological Pump in the Past
Figure 3 The nested cycles of marine organic carbon,
including (i) net primary production and heterotrophic
respiration in the surface ocean, (ii) export production
from the surface ocean and respiration in the deep sea
followed by upwelling or mixing of the respired CO2
back to the surface ocean (the biological pump, the
subject of this chapter), and (iii) burial of sedimentary
organic carbon and its release to the atmosphere by
weathering. The slight imbalance of each cycle fuels the
longer-timescale cycles. The carbon reservoirs shown
refer solely to the fraction affected by organic carbon
cycling; for instance, the deep-ocean value shown is of
CO2 produced by respiration, not the total dissolved
inorganic carbon reservoir.
is 180– 200 ppmv, or ,80 – 100 ppmv lower. CO2
is a greenhouse gas, and model calculations
suggest that its changes play a significant role in
the energetics of glacial/interglacial climate
change (Weaver et al., 1998; Webb et al., 1997).
However, we have not yet identified the cause of
these variations in CO2. The explanation of
glacial/interglacial cycles in CO2 is a major
motivation in the paleoceanographic and paleoclimatological communities, for three reasons:
first, these CO2 changes are the most certain of
any in our record of the past (excluding the
anthropogenic rise in CO2); second, they show a
remarkably strong connection to climate, inspiring
considerations of both cause and effect; and third,
they are of a timescale that is relevant to the
history and future of humanity.
The ultimate pacing of glacial cycles is
statistically linked to cyclic changes in the orbital
parameters of the Earth, with characteristic
frequencies of approximately 100 kyr, 41 kyr,
and 23 kyr (Berger et al., 1984; Hays et al.,
1976). These orbitally driven variations in the
seasonal and spatial distribution of solar radiation
incident on the Earth surface, known as the
“Milankovitch cycles”, are thought by many to
be the fundamental driver of glacial/interglacial
oscillations in climate. However, positive feedbacks within the Earth’s climate system must
amplify orbital forcing to produce the amplitude
and temporal structure of glacial/interglacial
climate variations. Carbon dioxide represents
one such feedback. The point of greatest uncertainty in this feedback is the mechanism by which
atmospheric CO2 is driven to change.
Broecker (1982a,b) first hypothesized a glacial
increase in the strength of the biological pump as
the driver of lower glacial CO2 levels. This
suggestion has spawned many variants, which
we will tend to phrase in terms of the “major”
nutrients, nitrogen and phosphorus. These nutrients are relatively scarce in surface ocean waters
but are required in large quantities to build all
algal biomass. In terms of the major nutrients, the
biological pump hypotheses fall into two groups:
(i) changes in the low and mid-latitude surface
ocean, where the major nutrients appear to limit
the extraction of CO2 by biological production,
and (ii) changes in the polar and subpolar ocean
regions, where the major nutrients are currently
not completely consumed and not limiting. In both
cases, the central biological process is “export
production,” the organic matter produced by
phytoplankton that is exported from the surface
ocean, resulting in the sequestration of its
degradation products (inorganic carbon and
nutrients) in the ocean interior. Here, we review
the concepts, tools, and observations relevant
to variability in the biological pump on the
millenial timescale, with a strong focus on its
potential to explain glacial/interglacial CO2
change.
The biogenic rain to the deep sea has important
mineral components: calcium carbonate, mostly
from coccolithophorids (phytoplankton) and foraminifera (zooplankton), and opal, mostly from
diatoms (phytoplankton) and radiolaria (zooplankton). The calcium carbonate (CaCO3) component is important for atmospheric carbon
dioxide in its own right (Figure 2). In contrast to
organic carbon, the production, sinking, and burial
of CaCO3 acts to raise atmospheric carbon dioxide
concentrations. This is unintuitive, in that CaCO3,
like organic carbon, represents a repository for
inorganic carbon and a vector for the removal of
Introduction
0
50
100
150
200
250
495
300
350
400
250
CO2
200
–400
dD
–450
dD (‰)
Atmospheric CO2 (ppm)
300
Ice sheet growth
–500
Ice volume
(relative,
inverted)
0
50
100
150
200
kyr BP
250
300
350
400
Figure 4 The history of atmospheric CO2 back to 400 kyr as recorded by the gas content in the Vostok ice core from
Antarctica (Petit et al., 1999). The ratio of deuterium to hydrogen in ice (expressed as the term dD) provides a record
of air temperature over Antarctica, with more negative dD values corresponding to colder conditions. The history of
global ice volume based on benthic foraminiferal oxygen isotope data from deep-sea sediment cores (Bassinot et al.,
1994) is plotted as relative sea level, so that ice ages (peaks in continental ice volume) appear as sea-level minima,
with a full glacial/interglacial amplitude for sea level change of ,130 m (Fairbanks, 1989). During peak glacial
periods, atmospheric CO2 is 80 – 100 ppmv lower than during peak interglacial periods, with upper and lower limits
that are reproduced in each of the 100 kyr cycles. Ice cores records, including the Vostok record shown here, indicate
that atmospheric CO2 was among the first parameters to change at the termination of glacial maxima, roughly in step
with southern hemisphere warming and preceding the decline in northern hemisphere ice volume.
this carbon from the surface ocean and atmosphere. The difference involves ocean “alkalinity.”
Alkalinity is the acid-titrating capacity of the
ocean. As it increases, the pH of seawater rises
(i.e., the concentration of H þ, or protons,
decreases), and an increasing amount of CO2 is
stored in the ionic forms bicarbonate (HCO2
3 ) and
carbonate (CO22
).
This
“storage”
is
achieved
by
3
the loss of Hþ from carbonic acid (H2CO3), which
is itself formed by the combination of dissolved
CO2 with its host H2O. Thus, an increase in ocean
alkalinity lowers the pCO2 of surface waters and
thus the CO2 concentration of the overlying
atmosphere. The carbonate ion holds two equivalents of alkalinity for every mole of carbon: CO22
3
must absorb two protons before it can leave the
ocean as CO2 gas. In the precipitation of CaCO3,
CO22
is removed from the ocean, lowering the
3
alkalinity of the ocean water and thus raising its
pCO2. We can describe this in terms of a chain of
reactions: the CO22
3 that was lost to precipitation
is replaced by the deprotonation of a HCO2
3,
generating a proton. This proton then combines
with another HCO2
3 to produce H2CO3, which
dissociates to form CO2 and H2O, yielding the
summary reaction: Ca2þ þ 2HCO2
3 ! CaCO3 þ
CO2 þ H2O. Thus, the biological precipitation of
CaCO3 raises the pCO2 of the water in which it
occurs.
The biological formation of CaCO3 affects
atmospheric CO2 from two perspectives. First,
the precipitation of CaCO3 in surface waters and
its sinking to the seafloor drives a surface-to-deep
alkalinity gradient, which raises the pCO2 of
surface waters in a way that is analogous to the
pCO2 decrease due to the biological pump; this
might be referred to as the “carbonate pump.” Just
as with the CO2 produced in deep water by organic
matter degradation, the chemical products of the
deep redissolution of the CaCO3 rain are eventually mixed up to the surface again, undoing the
effect of their temporary (,1,000 yr) sequestration in the abyss. Second, ,25% of the CaCO3
escapes dissolution and is buried, thus sequestering carbon and alkalinity in the geosphere, on a
timescale of thousands to millions of years
(Figure 2). An excess in the loss of alkalinity by
calcium carbonate burial rate relative to the input
of alkalinity by continental weathering will drive
an increase in the pCO2 of the whole ocean on the
496
The Biological Pump in the Past
timescale of thousands of years. In summary,
CaCO3 precipitation can alter atmospheric pCO2 by
generating a surface-to-deep gradient in
seawater alkalinity (the carbonate pump) and by
changing the total amount of alkalinity in the
ocean.
The CaCO3 cycle is a central part of the
effect of biological productivity on atmospheric
CO2. However, it is not within the strict definition
of the biological pump, which deals specifically
with organic carbon. Moreover, the effect of the
CaCO3 rain is determined not only by the
actual magnitude of the rain to the seafloor but
also by its degree of preservation and burial at the
seafloor, a relatively involved subject that is
treated elsewhere in this volume. Thus, in our
discussions below, we try as much as possible to
limit ourselves to the geochemical effects of
the biogenic rain of organic matter, bringing
CaCO3 into the discussion only when absolutely
necessary and then trying to focus on its
biological production and not its seafloor
preservation.
6.18.2 CONCEPTS
At the coarsest scale, phytoplankton abundance
(Figure 5(a)) and nutrient availability (Figures
5(b) –(d)) are correlated across the global surface
ocean, indicating that the supply of major
nutrients is a dominant control on productivity.
In this context, two types of environments emerge
from the global ocean distributions (Figure 5):
(i) the tropical and subtropical ocean, where
productivity is low and limited by the supply of
nitrate and phosphate, and (ii) the subpolar and
polar regions, where the supply of nitrate and
phosphate is high enough so as not to limit
phytoplankton growth. This distinction frames
much of the conceptual discussion that follows.
One important result will at this point seem
counterintuitive: the global biological pump is
driven from the low-latitude regions, where the
biological pump is working at maximal efficiency
with respect to the major nutrient supply, and
weakened by the high-latitude regions, which
are not maximally efficient with respect to the
major nutrients and which are thus allowing
deeply sequestered CO2 to escape back to the
atmosphere.
The relative importance of the low versus high
latitudes in controlling atmospheric CO2 is
currently a matter of debate, with the lines
drawn between the two types of models that are
used to study the biological pump: geochemical
box models of the ocean, the first tools available
for quantifying the role of the biological pump,
and the newer ocean general circulation models.
Relative to box models, at least some general
circulation models express a greater sensitivity of
atmospheric CO2 to the low-latitude ocean and a
lesser sensitivity to the high-latitude ocean
(Archer et al., 2000b). This difference in sensitivity has major implications for the feasibility of
different oceanic changes to explain glacial/
interglacial CO2 change. While some authors
have argued that the discrepancy can be framed in
terms of the amount of vertical mixing in the low
versus high-latitude ocean (Broecker et al., 1999),
we are not convinced by this argument, as such
differences should be captured by the deep-ocean
temperatures in the two types of models. Rather,
we favor the arguments of Toggweiler et al.
(2003a) that the difference is mostly in the degree
of air/sea equilibration of CO2 that is allowed in
the high-latitude regions of the models. If this is
so, there is good reason to believe that the high
latitudes are as important in the biological pump
as the box models suggest (Toggweiler et al.,
2003a). Regardless of our view, it must be
recognized that the quantitative arguments made
below are based largely on our experience with
box models (Hughen et al., 1997; Sigman et al.,
1998, 1999b, 2003), and that some workers would
find different areas of emphasis, although the same
basic processes would probably be described (e.g.,
see Archer et al., 2000a).
While the global correlation between nutrients
and productivity is strong, it is far from perfect.
For instance, the chlorophyll levels among the
high-latitude regions do not correlate with their
surface nutrient concentration, nor does such a
correlation hold particularly well within a single
high-latitude region (Figure 5). These discrepancies provide a starting point for addressing what
controls polar productivity and how the different
polar regions impact the biological pump. We also
look to the current differences among the polar
regions to understand how the high-latitude leaks
in the biological pump might have changed in
the past.
The general link between nutrient concentration and productivity is also broken in some
low-latitude regions. For instance, coastal productivity is high even where there is no apparent
elevation in major nutrients. While part of this
discrepancy is probably due to nutrient inputs
from the continents that are not captured in
Figure 5 because they are sporadic and/or
compositionally complex, the discrepancy also
reflects the very rapid cycling of organic matter
in the shallow ocean. We will see that the coastal
regions, despite their extremely high productivity, are not central to the biological
pump, illustrating the distinction between productivity and the biological pump.
While this comparison between “nutrients” and
productivity is useful, the surface distributions of
the nutrients are themselves not all identical.
Concepts
497
Figure 5 The concentrations in the surface ocean of chlorophyll (a qualitative index of phytoplankton abundance
2
and thus net primary productivity, (a)), the “major nutrients” phosphate (PO32
4 , (b)) and nitrate (NO3 , (c)), and
42
the nutrient silicate (SiO4 , (d)), which is required mostly by diatoms, a group of phytoplankton that builds opal tests.
The chlorophyll map in (a) is from composite data from the NASA SeaWiFs satellite project collected during 2001
(http://seawifs.gsfc.nasa.gov/SEAWIFS.html). The nutrient data in (b), (c), and (d) are from 1994 World Ocean Atlas
(Conkright et al., 1994), and the maps are generated from the International Research Institute for Climate Prediction
Data Library (http://ingrid.ldeo.columbia.edu). Comparison of (a) with (b), (c), and (d) demonstrates that
phytoplankton abundance, at the coarsest scale, is driven by the availability of these nutrients, and these nutrients are
generally most available in the polar ocean and along the equatorial upwellings where nutrient-rich deep water
mixes more easily to the surface. However, among the nutrient-bearing high-latitude regions, there is not a good
correlation between nutrient concentration and chlorophyll, indicating that other parameters come to limit
productivity in these regions. The high-latitude ocean of the southern hemisphere, which has the highest nutrient
concentrations, is known as the Southern Ocean. It is composed of the more polar Antarctic zone, where the silicate
concentration is high and diatom productivity is extensive, and the more equatorward Subantarctic zone, where
nitrate and phosphate remain high but silicate is scarce. The strong global correlation between phosphate and nitrate,
originally recognized by Redfield (Redfield, 1942, 1958; Redfield et al., 1963), allows us to group the two nutrients
together when considering their internal cycling within the ocean, as they are exhausted in more or less the same
regions. However, their input/output budgets may change over hundreds and thousands of years, and the relationship
between these two nutrients may have been different in the past.
498
The Biological Pump in the Past
Figure 5
For instance, nitrate and phosphate penetrate
much further north than does silicate in the
Southern Ocean, the polar ocean surrounding
Antarctica (Figures 5(b) – (d)). As silicate is
needed by diatoms but not other types of
phytoplankton, this difference has major implications for the ecology of the Southern Ocean,
present and past. Moreover, since surface water
from 40 –508 S is fed into the thermocline that
(continued).
supplies the low-latitude surface ocean with
nutrients, the nutrient conditions of the Southern
Ocean affect the nutrient supply and ecology
of lower latitudes (Brzezinski et al., 2002;
Matsumoto et al., 2002b). With regard to nitrate
and phosphate (Figures 5(b) and (c)), a more
careful comparison of their concentration variations would show lower nitrate-to-phosphate
ratios in the Pacific than elsewhere, indicative of
Concepts
the nitrate loss that is occurring in that basin
(Deutsch et al., 2001; Gruber and Sarmiento,
1997). The input/output budget of the marine
nitrogen cycle and its affect on the ocean nitrate
reservoir arises as a matter of great importance in
the section that follows.
Equation (2) for [DIC]deep 2 [DIC]surface gives
½DICdeep 2 ½DICsurface ¼ ðEPP £ C=PEP Þ=Q
ð3Þ
Solving Equation (1) for EPP gives:
32
EPP ¼ ð½PO32
4 deep 2 ½PO4 surface Þ £ Q
ð4Þ
Substitution of Equation (4) into Equation (3) for
EPP gives the following expression for
[DIC]deep 2 [DIC]surface :
6.18.2.1 Low- and Mid-latitude Ocean
We can describe the global biological pump in
terms of a balance of fluxes between the sunlit
surface ocean and the cold, dark, voluminous deep
ocean (Figure 1). Phytoplankton consume CO2
and nutrients (in particular, the major nutrients
nitrogen and phosphorus) in the surface, and
subsequent processing drives a downward rain of
this organic matter that is degraded after it has
sunk into the ocean interior, effectively extracting
the CO2 and nutrients from the surface ocean and
lowering atmospheric CO2. Balancing this downward flux of particles is the net upward flux of
dissolved inorganic carbon and nutrients into the
surface, due to a combination of diffuse vertical
mixing over the entire ocean and focused vertical
motion (upwelling and downwelling) in specific
regions.
This steady state can be described by the
following expressions (Broecker and Peng,
1982), which indicate a balance between the
gross upward (left-hand side) and gross downward
(right-hand side) transport of phosphorus
(Equation (1)) and carbon (Equation (2)):
32
½PO32
4 deep £ Q ¼ ½PO4 surface £ Q þ EPP
499
ð1Þ
½DICdeep £ Q ¼ ½DICsurface £ Q þ ðEPP £ C=PEP Þ
ð2Þ
where [PO32
4 ] is the phosphate concentration and
[DIC] is the dissolved inorganic carbon concentration (the sum of dissolved CO2, HCO2
3 , and
CO22
3 ) in the deep and surface layers, Q is the
exchange rate of water between the deep and
surface, EPP is the export production in terms of
phosphorus, and C/PEP is the carbon to phosphorus ratio of the export production. In these
expressions, phosphorus is intended to represent
the major nutrients in general; the distinctions
between phosphorus and nitrogen will be considered later.
The biological pump is the biologically driven
gradient in [DIC] between the surface and the deep
sea ([DIC]deep 2 [DIC]surface). Because the deep
ocean is a very large reservoir, increasing this
gradient essentially means decreasing [DIC]surface.
Because the surface ocean equilibrates with the
gases of the atmosphere, an increase in the
gradient lowers atmospheric CO2 as well. Solving
½DICdeep 2 ½DICsurface
32
¼ ð½PO32
4 deep 2 ½PO4 surface Þ £ C=PEP
ð5Þ
Thus, given the assumptions above, the surface–
deep gradient in inorganic carbon concentration (the
“strength” of the biological pump) is determined by
(i) the C/P of the organic matter that is exported from
the surface ocean, and (ii) the nutrient concentration
([PO32
4 ]) gradient between the surface and deep
ocean.
In the low- and mid-latitude ocean, this balance
is simplified by the nearly complete extraction of
2
[PO32
4 ] and [NO 3 ] from the surface ocean
(Figures 5(b) and (c)). The warm sunlit surface
layer is separated by a strong temperature-driven
density gradient (“thermocline”) from the cold
deep ocean, preventing rapid communication
between the low-latitude ocean surface and the
ocean interior. As a result, nutrient supply from
below occurs slowly, and phytoplankton completely strip surface waters of available nitrogen and
phosphorus; i.e., low- and mid-latitude phytoplankton growth is limited by the major nutrients
nitrogen and phosphorus. If the low-latitude
surface ocean is so strongly limited by the major
nutrients that [PO32
4 ]surface would not be raised
above zero for any imaginable increase in Q
(i.e., if surface ocean plant productivity remains
major-nutrient limited over a broad range of
surface – deep mixing rates), then changes in
upwelling and vertical mixing are balanced by
equivalent changes in biological export production. In this case, changes in Q could cause
changes in productivity over time, but these would
not change the surface –deep gradient in [DIC]
maintained by the biological pump.
If this is an adequate description of the real lowlatitude ocean, then the strength of its biological
pump (the amplitude of the surface – deep gradient
in inorganic carbon concentration) is controlled
solely by: (i) the carbon/nutrient ratio of the
organic matter, and (ii) the nutrient content of
the deep ocean (see Equation (5)). If the C/P ratio
of export production increased or the [PO32
4 ] of
the deep ocean increased, then biological production at the surface would drive an increase in
the downward flux of carbon that is not at first
matched by any change in the upward flux of
carbon dioxide associated with surface – deep
500
The Biological Pump in the Past
mixing. A new balance between the downward
and upward fluxes of carbon would eventually be
reached as the [DIC] gradient between the surface
and the deep increases, so that a given amount of
mixing transports more CO2 back into the surface
ocean. This increase in surface – deep [DIC]
gradient would be mostly due to a drop in
dissolved inorganic carbon concentration of the
surface ocean, which would lower the pCO2 of the
surface ocean and drive an associated decrease in
atmospheric CO2.
Biological production in the ocean tends to
incorporate carbon, nitrogen, and phosphorus into
biomass in the ratios of 106/16/1 (Redfield, 1942,
1958; Redfield et al., 1963). These ratios determine the amount of inorganic carbon that is
sequestered in the deep sea when the supply of
nutrients to the surface is completely converted to
export production, as is the case in the modern
low- and mid-latitude ocean. It is unclear at this
point how variable the C/N/P ratios of export
production might be through time. When
Broecker (1982a) first hypothesized a glacial
increase in the strength of the biological pump
as the driver of lower glacial CO2 levels, he
considered the possibility (suggested to him by
Peter Weyl) that the C/P export production
changes between ice ages and interglacials, with
a higher C/P ratio during the last ice age, but he
could find neither a mechanism nor direct
paleoceanographic evidence for such a change.
Roughly the same situation persists today. We
remain exceedingly uncertain about the robustness
of the “Redfield ratios” through time and space.
However, we do not yet have any mechanistic
understanding of their variability (Falkowski,
2000), neither have we recognized a paleoceanographic archive that would help us to assess this
variability. The organic matter preserved in the
sediment record represents a minute and compositionally distinct fraction of the organic matter
exported from the surface ocean. We have no
reason to believe that variations in the elemental
ratios of this residual sedimentary organic matter
will reflect that of export production, especially in
the face of changing seafloor conditions.
One of the persistent views of geochemists, to
the frustration of their more biologically oriented
colleagues, is that the complex recycling of
elements in the low-latitude surface ocean has
no major significance for the low-latitude biological pump. The argument is that, given the nutrient
limitation that predominates in the low-latitude
surface, the amount of carbon to be exported is
determined by the supply of the major nutrients
from the subsurface and the C/N/P uptake ratios of
phytoplankton, regardless of the upper ocean
recycling of a given nutrient element before
export. However, the C/N/P ratios of export
production may be altered by ecological effects
on the relationship between phytoplankton biomass and the organic matter actually exported
from the surface ocean, even without changes in
the elemental requirements of the phytoplankton
themselves. As a result, the ecology of the surface
ocean could drive variations in the strength
of the biological pump, for instance, by changing
the C/P ratio of export production independent of
the C/P ratio of phytoplankton, so as to change the
amount of carbon exported for a given phosphorus
supply. It was with this reasoning that Weyl
originally justified the idea of glacial/interglacial
changes in the C/P ratio of export production
(Broecker, 1982a). Terrestrial biogeochemists and
limnologists are actively seeking a quantitative
theory for the ecological constraints imposed by
elemental cycling (Elser et al., 2000); oceanographers may benefit from a similar focus on the
controls on elemental ratios, in particular, in
export production. In any case, while changes in
the chemical composition of export production
may play a role in the variability of the ocean’s
biological pump, we are not yet in a position to
posit specific hypotheses about such changes or to
recognize them in the sedimentary record.
Broecker (1982b) and McElroy (1983) initially
hypothesized that the oceanic reservoirs of
phosphate and nitrate, respectively, might
increase during glacial times, which would allow
enhanced low-latitude biological production to
lower atmospheric CO2. However, it was recognized that this basic mechanism would require
large changes in nutrient reservoirs to produce the
entire observed amplitude of CO 2 change.
According to a box-model calculation, the
immediate effect of an increase in the nutrient
reservoir is to extract CO2 from the surface ocean
and atmosphere, sequestering it in the deep sea,
with a 30% increase in the oceanic nutrient
reservoir driving a 30 –40 ppmv CO2 decrease
(Table 1, see “closed system effects”) (Sigman
et al., 1998). The longer-term effect on CO2 from
the ocean alkalinity balance depends greatly on
whether CaCO3 export increases in step with Corg
export. It is not known whether a global increase
in low-latitude production would increase the
CaCO3 rain in step with the Corg rain—this
would probably depend on the circumstances of
Table 1 Atmospheric CO2 effects of a 30% increase in
the ocean nutrient inventory.
CO2 changes
(ppm)
CaCO3 production
Proportional to Corg Constant
Closed system effects
Open system effects
234
þ17
243
23
Total CO2 change
217
246
Concepts
the increase—so we should consider both
possibilities.
If CaCO3 export increases proportionately with
Corg export (see “CaCO3 production proportional
to Corg” in Table 1), the increase in the CaCO3 flux
to the seafloor causes a gradual loss of alkalinity
from the ocean, which works to raise atmospheric
CO2 (see “open system effects” in Table 1). As a
result, only a 15– 25 ppmv drop in CO2 results
from a 30% increase in the oceanic major nutrient
reservoir (Table 1). However, if CaCO3 production remains constant as Corg production
increases (see “CaCO3 production constant” in
Table 1), the effects on the whole ocean reservoir
of alkalinity are minimal (see “open system
effects” in Table 1), and the increase in the
oceanic nutrient reservoir is much more effective
at lowering atmospheric CO2. In this case, the box
model calculation suggests that a 30% increase
in the oceanic nutrient reservoir lowers atmospheric CO2 by ,50 ppmv (Table 1), while some
general circulation models would suggest declines
in CO2 of as much as 80 ppm (Archer et al.,
2000a). The required increase in the ocean
nutrient reservoir could perhaps be lowered by a
coincident decrease in CaCO3 export that was
modest enough not to violate observations of the
glacial-age sedimentary CaCO 3 distribution
(Archer and Maier-Reimer, 1994; Brzezinski
et al., 2002; Matsumoto et al., 2002b; Sigman
et al., 1998). Nevertheless, a 30% increase in
oceanic nutrients would require a dramatic change
in ocean biogeochemistry.
Given a residence time of .6 thousand years
for oceanic phosphate and the nature of the
input/output terms in the oceanic phosphorus
budget (Froelich et al., 1982; Ruttenberg, 1993),
it is difficult to imagine how oceanic phosphate
could vary so as to cause the observed amplitude,
rate, and phasing of atmospheric CO2 change
(Ganeshram et al., 2002). Entering an ice age,
Broecker (1982a) considered the possibility of a
large phosphorus input from the weathering of
shelf sediments, as ocean water is captured in ice
sheets and sea level drops. Ice-core studies
dispatched with this admittedly unlikely scenario
by demonstrating that, going into an interglacial
period, CO2 begins to rise before sea level rises
(Sowers and Bender, 1995), so that the phosphorus inventory decrease would occur too late to
explain the rise of CO2 into interglacials.
There is growing evidence that the nitrogen
cycle is adequately dynamic to allow for a large
change in the oceanic nitrate reservoir on
glacial/interglacial timescales (Codispoti, 1995;
Gruber and Sarmiento, 1997). Denitrification, the
heterotrophic reduction of nitrate to N2 gas, is the
dominant loss term in the oceanic budget of
“fixed” (or bioavailable) nitrogen. Nitrogen isotope studies in currently active regions of
501
denitrification provide compelling evidence that
water column denitrification was reduced in these
regions during glacial periods (Altabet et al.,
1995; Ganeshram et al., 1995, 2002; Pride et al.,
1999). In addition, it has been suggested that N2
fixation, the synthesis of new fixed nitrogen from
N2 by cyanobacterial phytoplankton, was greater
during glacial periods because of increased
atmospheric supply of iron to the open ocean
(iron being a central requirement of the enzymes
for N2 fixation) (Falkowski, 1997). Both of these
changes, a decrease in water column denitrification and an increase in N2 fixation, would have
increased the oceanic nitrate reservoir. It has been
suggested that such a nitrate reservoir increase
would lead to significantly higher export production in the open ocean during glacial periods,
potentially explaining glacial/interglacial atmospheric CO2 changes (Broecker and Henderson,
1998; Falkowski, 1997; McElroy, 1983).
The major conceptual questions associated with
this hypothesis involve the feedback of the
nitrogen cycle and the strictness with which
marine organisms must adhere to the Redfield
ratios, both N/P and C/N. Oceanic N2 fixation
provides a mechanism by which the phytoplankton community can add fixed nitrogen to the ocean
when conditions favor it. By contrast, riverine
input of phosphorus, the major source of phosphate to the ocean, is not directly controlled by
marine phytoplankton. Because phytoplankton
have no way to produce biologically available
phosphorus when there is none available, geochemists have traditionally considered phosphorus to
be the fundamentally limiting major nutrient on
glacial/interglacial timescales (Broecker and
Peng, 1982). For a nitrate reservoir increase
alone to drive an increase in low-latitude biological production, the nutrient requirements of the
upper-ocean biota must be able to deviate from
Redfield stoichiometry to adjust to changes in the
N/P ratio of nutrient supply, so as to fully utilize
the added nitrate in the absence of added
phosphate. Alternatively, if this compensatory
shift in biomass composition is incomplete, the
surface ocean will tend to shift toward phosphate
limitation, and an increase in export production
will be prevented.
It was originally suggested by Redfield
(Redfield et al., 1963), based on analogy with
freshwater systems, that N2 fixers in the open
ocean will enjoy greater competitive success
under conditions of nitrate limitation but will be
discouraged in the case of phosphate limitation.
Having posited this sensitivity, Redfield hypothesized that N2 fixation acts as a negative feedback
on the nitrate reservoir, varying so as to prevent
large changes in the nitrate reservoir that are not
associated with a coincident change in the
phosphate reservoir (Redfield et al., 1963).
502
The Biological Pump in the Past
On the grand scale, this feedback must exist and
must put some bounds on the degree to which the
oceanic nitrate reservoir can vary independently
from phosphorus. However, the quantitative constraint that this negative feedback places on the
global nitrate reservoir is not yet known, with one
suggested possibility being that iron is so
important to N2 fixers that changes in its
supply can overpower the nitrate/phosphate sensitivity described by Redfield, leading to significant variations in the nitrate reservoir over
time (Broecker and Henderson, 1998; Falkowski,
1997).
In summary, there are two components of the
paradigm of phosphorus’s control on the nitrate
reservoir and the low-latitude biological pump,
both stemming from Redfield’s work: (i) oceanic
N2 fixation proceeds in phosphorus-bearing,
nitrogen-poor environments, but not otherwise,
and (ii) export production has a consistent C/N/P
stoichiometry. The first statement would have to
be inaccurate for the nitrate reservoir to vary
independently of the phosphate reservoir, and the
second statement would have to be inaccurate for
such an independent change in the nitrate
reservoir to actually cause a change in the
biological pump. Partly as a reaction to studies
that have considered the influence of other
parameters, in particular, the supply of iron to
open ocean N2 fixation (Falkowski, 1997), some
papers have restated the traditional Redfield-based
paradigm (Ganeshram et al., 2002; Tyrrell, 1999).
However, the real challenge before us is to test and
quantify the rigidity of the Redfield constraints
relative to the influence of other environmental
parameters. Biogeochemical studies of the modern ocean (Karl et al., 1997; Sanudo-Wilhelmy
et al., 2001) and the paleoceanographic record
(Haug et al., 1998) are both likely to play a role in
the evaluation of these questions regarding Redfield stoichiometry and N2 fixation, which are, in
turn, critical for the hypothesis of the low latitude
biological pump as the driver of glacial/interglacial CO2 change.
The mean concentration of O2 in the ocean
interior would be lowered in the glacial ocean by
an increase in low-latitude carbon export, regardless of whether it is driven by an increase in the
ocean phosphorus reservoir, the ocean nitrogen
reservoir, or a change in the carbon/nutrient ratio
of sinking organic matter. As described below,
this makes deep-ocean [O2] a possible constraint
on the strength and efficiency of the biological
pump in the past. Here, however, we focus on this
sensitivity of deep-ocean [O2] as part of an
additional negative feedback in the ocean
nitrogen cycle. This feedback may restrict the
variability of the oceanic nitrate reservoir, even
in the case that Redfield’s N2 fixation-based
feedback on the ocean nitrogen cycle is weak
and ineffective.
Denitrification occurs in environments that are
deficient in O2. An increase in the ocean’s nitrate
content will drive higher export production
(neglecting, for the moment, the possibility of
phosphate limitation). When this increased export
production is oxidized in the ocean interior, it will
drive more extensive O2 deficiency. This will lead
to a higher global rate of denitrification, which, in
turn, will lower the ocean’s nitrate content. Thus,
the sensitivity of denitrification to the O2 content
of the ocean interior generates a hypothetical
negative feedback that may, like the N2 fixationbased feedback, work to stabilize the nitrate
content of the ocean (Toggweiler and Carson,
1995).
While it is generally true that the major
nutrients nitrate and phosphate are absent in
low- and mid-latitude surface ocean, there are
important exceptions (Figures 5(b) and (c)).
Wind-driven upwelling leads to nonzero nutrient
concentrations and high biological productivity at
the surface along the equator and the eastern
margins of the ocean basins (e.g., off of Peru,
California, and western Africa). Why have we
given such short shrift to these biologically
dynamic areas in our discussion of the low- and
mid-latitude biological pump?
While critically important for ocean ecosystems
and potentially important for interannual variations in atmospheric CO2, the nutrient status in
equatorial and coastal upwelling systems does not
greatly affect atmospheric CO2 on centennial or
millennial timescales. Above, we demonstrated
that greater vertical exchange between the
nutrient-poor surface ocean and the deep sea
causes higher export production but does not
affect the [DIC] gradient between the surface and
the deep sea, so that it would not drive a change in
atmospheric CO2. Following this same reasoning,
low-latitude upwellings may generate much
export production, but they are not millennialscale sinks for atmospheric CO2 because the
upwelling brings up both nutrients and respiratory
CO2 from the ocean interior. Neither do the
upwellings drive a net loss of CO2 from the ocean,
because the high-nutrient surface waters of the
low-latitude upwellings do not contribute appreciably to the ventilation of the ocean subsurface. Put
another way, the nutrients supplied to the surface
are eventually consumed as the nutrient-bearing
surface water flows away from the site of
upwelling, before the water has an opportunity
to descend back into the ocean interior. This
pattern is evident in surface ocean pCO2 data
(Takahashi et al., 1997). For instance, in the core
of equatorial Pacific upwelling, pCO2 is high both
because the water is warming and because
deeply sequestered metabolic CO2 is brought to
Concepts
the surface and evades back to the atmosphere.
Adjacent to the equatorial upwelling, the tropical
and subtropical Pacific is a region of low pCO2
which is associated with the consumption of the
excess surface nutrients that have escaped consumption at the site of upwelling. It does not
matter for atmospheric CO2 whether the upwelled
nutrients are converted to export production onaxis of the upwelling system or further off-axis.
Thus, while changes in upwelling, biological
production, and nutrient status are central to the
history of the low-latitude ocean, they have
limited importance for atmospheric CO2 changes
on glacial/interglacial timescales. Increased
export production driven by higher rates of
vertical mixing or upwelling would only play a
role in lowering atmospheric CO2 if it caused a
change in the chemical composition of the
exported organic material, such as a decrease in
its CaCO3/Corg ratio or an increase in the mean
depth at which the organic rain is metabolized in
the ocean interior (e.g., Boyle, 1988b; Dymond
and Lyle, 1985).
The surface chlorophyll distribution shows that
coastal environments, even in regions without
upwelling, are among the most highly productive
in the ocean (Figure 5(a)). However, it is believed
that coastal ocean productivity does not have a
dominant effect on atmospheric CO2 over centuries and millennia. As biomass does not
accumulate significantly in the ocean, any longterm excess in net primary production relative to
heterotrophy must lead to export of organic matter
out of the surface layer, in particular, as a
downward rain of biogenic particles. In the case
of the coastal ocean, this export is stopped by the
shallow seafloor, frequently within the depth
range that mixes actively with the surface ocean.
The high net primary productivity of the coastal
ocean owes much to the presence of the shallow
seafloor, as the nutrients released from the
degradation of settled organic matter are available
to the phytoplankton in the sunlit surface. Yet, just
as the seafloor prevents the loss of nutrients from
the coastal surface ocean, so too does it prevent
the export of carbon. Almost as soon as the settled
organic carbon is respired, the CO2 product is free
to diffuse back into the atmosphere. Were the
same export to have occurred over the open ocean,
it would have effectively sequestered the carbon in
that biogenic rain within the ocean interior for
roughly a thousand years. Thus, the shallow
seafloor of the ocean margin leads to high primary
productivity while at the same time limiting its
effect on the carbon cycle over the timescale of
centuries and millennia. It should be noted,
however, that the ocean margins are critically
important in the global carbon cycle on the
timescale of millions of years, due to their role
in organic carbon burial.
503
It has been hypothesized that a significant
fraction of the organic matter raining onto
the shallow margin is swept over the shelf/slope
break, thus transporting organic matter into the
ocean interior (Walsh et al., 1981). This has been
interpreted as rendering ocean margin productivity important to atmospheric CO2 on
millennial timescales, in that the organic matter
being swept off the shelf would be an important
route by which CO2 is shuttled into the deep sea.
However, this view should be considered carefully. The ocean margins have high net primary
production because of nutrient recycling, but the
recycled nutrients were produced by the
decomposition of organic matter that would have
otherwise been exported out of the system. Mass
balance dictates that export of organic carbon
from the coastal surface ocean, like export
production in the open ocean surface, is set by
the net supply of nutrients to the system. If a lowlatitude ocean margin receives most of its net
nutrient supply from the deep sea, then it is no
different than the neighboring open ocean in the
quantity of carbon that it can export. There is one
caveat to this argument: if organic matter
deposited on the shelf can have its nutrients
stripped out without oxidizing the carbon back to
CO2 and this “depleted” organic carbon is then
transported off the shelf and into the ocean
interior, then ocean margin productivity, having
been freed from the Redfield constraint on the
elemental composition of its export production,
could drive a greater amount of organic carbon
export for a given amount of nutrient supply.
6.18.2.2 High-latitude Ocean
Broecker’s (Broecker, 1981, 1982a,b; Broecker
and Peng, 1982) initial ideas about the biological
pump revolutionized the field of chemical oceanography. However, soon after their description,
several groups demonstrated that his focus on the
low-latitude ocean missed important aspects of the
ocean carbon cycle. The thermocline outcrops at
subpolar latitudes, allowing deep waters more
ready access to the surface in the polar regions. In
these regions, the nutrient-rich and CO2-charged
waters of the deep ocean are exposed to the
atmosphere and returned to the subsurface before
the available nutrients are fully utilized by
phytoplankton for carbon fixation. This incomplete utilization of the major nutrients allows for
the leakage of deeply sequestered CO2 back into
the atmosphere, raising the atmospheric pCO2.
Work is ongoing to understand what limits
phytoplankton growth in these high-latitude
regions. Both light and trace metals such as iron
are scarce commodities in these regions and
together probably represent the dominant controls
504
The Biological Pump in the Past
on polar productivity, with light increasing in
importance toward the poles due to the combined
effects of low irradiance, sea ice coverage, and
deep vertical mixing (see Chapters 6.02, 6.04, and
6.10). The Southern Ocean, which surrounds
Antarctica, holds the largest amount of unused
surface nutrients (Figures 5(b) – (d)), yet the
surface chlorophyll suggests that it is perhaps
the least productive of the polar oceans; iron
and light probably both play a role in explaining
this pattern (Martin et al., 1990; Mitchell
et al., 1991).
The efficiency of the global biological pump
with respect to the ocean’s major nutrient content
is determined by the nutrient status of the polar
regions that ventilate the deep sea, or more
specifically, the nutrient concentration of the
new water that enters the deep sea from polar
regions (Figure 6). Deep water is enriched in
nutrients and CO2 because of the low- and midlatitude biological pump, which sequesters both
nutrients and inorganic carbon in subsurface.
In high-latitude regions such as the Antarctic,
the exposure of this CO2-rich deep water at the
surface releases this sequestered CO2 to the
atmosphere. However, the net uptake of nutrients
and CO2 in the formation of phytoplankton
biomass and the eventual export of organic matter
subsequently lowers the pCO2 of surface waters,
causing the surface layer to reabsorb a portion of
the CO2 that was originally lost from the upwelled
water. Thus, the ratio between export production
and the ventilation of CO2-rich subsurface water,
not the absolute rate of either process alone,
controls the exchange of CO2 between the
atmosphere and high-latitude surface ocean. The
nutrient concentration of water at the time that it
enters the subsurface is referred to as its
“preformed” nutrient concentration. The preformed nutrient concentration of subsurface
water provides an indicator of the cumulative
nutrient utilization that it underwent while at the
surface, with a higher preformed nutrient concentration indicating lower cumulative nutrient
utilization at the surface and thus a greater leak
in the biological carbon pump. This picture
overlooks a number of important facts. For
instance, ocean/atmosphere CO2 exchange is not
instantaneous, so water can sink before it has
reached CO2 saturation with respect to the
atmosphere. In polar regions, this typically
means that deep water may leave the surface
before it has lost as much CO2 to the atmosphere
as it might have (Stephens and Keeling, 2000).
The global efficiency of the biological pump
can be evaluated in terms of a global mean
preformed nutrient concentration for surface water
that is folded into the ocean interior. In the
calculation of this mean preformed nutrient
concentration, each subsurface water formation
term is weighted according to the volume of
the ocean that it ventilates because this corresponds to the amount of CO2 that could be sequestered in a given portion of the ocean interior. As a
result, the nutrient status of surface ocean regions
that directly ventilate the ocean subsurface (deep,
intermediate, or thermocline waters) have special
importance for atmospheric CO2.
The high-latitude North Atlantic and the
Southern Ocean are the two regions that appear
to dominate the ventilation of the modern deep
ocean. Through the preformed nutrient concentrations of the newly formed deep waters, their
competition to fill the deep ocean largely determines the net efficiency of the global biological
pump (Toggweiler et al., 2003b). Of these two
regions, the North Atlantic has a low preformed
nutrient concentration and thus drives the ocean
toward a high efficiency for the biological pump.
In contrast, the preformed nutrient concentration
of Southern Ocean-sourced deep water is high and
thus drives the global biological pump toward a
low efficiency.
Within the polar oceans involved in deep water
formation, certain regions are more important than
others. The Antarctic Zone, the most polar region
in the Southern Ocean, is involved in the
formation of both deep and intermediate-depth
waters, making this region important to the
atmosphere/ocean CO2 balance. The quantitative
effect of the Subantarctic Zone on atmospheric
CO2 is less certain, depending on the degree to
which the nutrient status of the Subantarctic
surface influences the preformed nutrient concentration of newly formed subsurface water
(Antarctic Intermediate Water and Subantarctic
mode water), but its significance is probably much
less than that of the Antarctic.
Questions regarding the most critical regions
extend to even smaller spatial scales. For instance,
with regard to deep-ocean ventilation in the
Antarctic, it is uncertain as to whether the entire
Antarctic surface plays an important role in the
CO2 balance, or whether instead only the specific
locations of active deep water formation (e.g., the
Weddell Sea shelf) are relevant. This question
hinges largely on whether the surface water in the
region of deep water formation exchanges with
the open Antarctic surface, or whether it instead
comes directly from the shallow subsurface and
remains isolated until sinking. If the latter is true,
then the open Antarctic may be irrelevant to the
chemistry of the newly formed deep water. This
would be problematic for paleoceanographers,
since essentially all of the reliable paleoceanographic records come from outside these rather
special environments. There is, however, a new
twist to this question. While deep water formation
in the modern Antarctic has long been thought of
as restricted to the shelves, efforts to reproduce
Concepts
505
Figure 6 The effect on atmospheric CO2 of the biological pump in a region of deep-ocean ventilation where the
major nutrients are not completely consumed, such as the Antarctic zone of the Southern Ocean. In this figure, NO2
3 is
intended to represent the major nutrients and thus should not be distinguished conceptually from PO32
4 . The
biological pump causes deep water to have dissolved inorganic carbon (DIC) in excess of its “preformed” [DIC] (the
concentration of DIC in the water when it left the surface; see lower left box, which represents water in the ocean
interior). The low- and mid-latitudes (and a few polar regions, such as the North Atlantic) house a biological pump
that is “efficient” in that it consumes all or most of the major nutrient supply. As a result, it drives the DIC excess
toward a concentration equivalent to the deep [NO2
3 ] multiplied by the C/N ratio of exported organic matter (left
lower arrow represents this influence on deep-water chemistry). The respiration of organic matter in the ocean interior
also leads deep water to have an O2 deficit relative to saturation. In high-latitude regions such as the Antarctic, the
exposure of this deep water at the surface releases this sequestered CO2 to the atmosphere and leads to the uptake of
O2 from the atmosphere (upper left). However, the net uptake of nutrients and CO2 in the formation of phytoplankton
biomass and the eventual export of organic matter (“uptake and export”) subsequently lowers the pCO2of surface
waters, causing the surface layer to reabsorb a portion of the CO2 that was originally lost from the upwelled water
(upper right). The net excess in photosynthesis to respiration that drives this export out of the surface also produces
O2, some of which evades to the atmosphere, partially offsetting the initial O2 uptake by the surface ocean. While the
down-going water has lost its DIC excess by exchange with the atmosphere, a portion of the original DIC excess will
be reintroduced in the subsurface when the exported organic matter is remineralized (the more complete the nutrient
consumption in the surface, the greater the DIC excess that is reintroduced in the subsurface). The ratio between
export and the ventilation of CO2-rich subsurface water, which controls the net flux of CO2 between the atmosphere
and surface ocean in this region, can be related to the nutrients. Nutrient utilization, defined as the ratio of the rate of
nutrient uptake (and export) to the rate of nutrient supply, expresses the difference between nutrient concentration of
rising and sinking water. Nutrient utilization relates directly (through the Redfield ratios) to the ratio of CO2 invasion
(driven by organic matter export) relative to evasion (driven by deep-water exposure at the surface). The lower the
nutrient utilization, the lower the ratio of invasion to evasion and thus the more this region represents a leak in the
biological pump. This diagram overlooks a number of important facts. For instance, ocean/atmosphere CO2 exchange
is not instantaneous, so water can sink before it has reached CO2 saturation with respect to the atmosphere.
deep-sea nutrient chemistry and radiochemistry
raise the possibility that the deep ocean may also
be ventilating in a more diffuse mode throughout
the open Antarctic (Broecker et al., 1998; Peacock,
2001). If this is the case, the significance of the
open Antarctic in the CO 2 balance is less
susceptible to uncertainties about lateral exchange
with the shelf sites of deep water formation. In the
discussion that follows, we assume that the open
Antarctic does affect the CO2 exchanges associated with deep water formation, either by active
exchange of surface water with the specific sites of
506
The Biological Pump in the Past
Figure 7 Schematic illustrations of the CO2 leak from the Southern Ocean to the atmospheric that exists today and
of two alternative scenarios by which this leak might have been stopped during ice ages. The incomplete consumption
of nutrients in the modern Southern Ocean causes waters that come to the surface in the Southern Ocean to release
CO2 to the atmosphere because not all of the nutrient supply to the surface drives a downward rain of organic
material, allowing CO2 sequestered by the lower latitude biological pump to escape (a). During ice ages, atmospheric
CO2 could have been reduced by increasing the nutrient efficiency of the biological pump in the Southern Ocean
(b and c). This could have been driven by an increase in biological productivity (b), which would have actively
increased the downward flux of carbon, or by density “stratificaton” of the upper ocean, allowing less nutrient- and
CO2-rich water to the Southern Ocean surface (c). As described in Figure 8, because of the existence of other regions
of deep water formation, the evasion of CO2 is reduced upon stratification (c) even if the degree of nutrient
consumption in the surface does not change.
deep water formation or by its actual participation
in the ventilation process. In any case, we must
improve our understanding of the modern physical
and chemical oceanography of the polar regions if
we are to determine their roles in atmospheric CO2
change.
In their hypothesis of a glacial decrease in
Southern Ocean CO2 leak, early workers considered two causes: (i) an increase in biological
export production and (ii) a decrease in the
exposure rate of deep waters at the polar surface
(Figure 7). One can imagine processes that would
have reduced the evasion of CO2 from the
Southern Ocean by either of the two mechanisms
mentioned above. For instance, an increase in the
input of dust and its associated trace metals to the
Southern Ocean might have driven an increase in
the rate of nutrient and carbon uptake by
phytoplankton (Martin, 1990). Alternatively, an
increase in the salinity-driven stratification of the
Antarctic and/or a decrease in wind-driven
upwelling could have lowered the rate of
nutrient and carbon dioxide supply to the surface
(discussed further in Section 6.18.3) (Francois
et al., 1997; Keeling and Visbeck, 2001; Sigman
and Boyle, 2000, 2001; Toggweiler and Samuels,
1995). Under most conditions, these two hypothesized changes do not have the same quantitative effect on atmospheric CO2. For instance,
stratification can cause a significantly larger
amount of CO2 decrease than does increased
export production for a given amount of nutrient
drawdown in Southern Ocean surface waters
(Sigman et al., 1999b).
These differences can be understood in the
context of the mean preformed nutrient concentration of the ocean interior (Figure 8). There are
two Southern Ocean mechanisms by which the
global efficiency of the biological pump may be
increased during glacial times: (i) decreasing the
preformed nutrient concentration of SouthernOcean-sourced deep water (Figures 8(c) and (d)),
and (ii) decreasing the importance of SouthernOcean-sourced deep waters in the ventilation of
the deep ocean, relative to the North Atlantic or
some other source of deep water with low
Concepts
507
Figure 8 Cartoons depicting the ventilation of the ocean interior by the North Atlantic and Southern Ocean and the
preformed [PO32
4 ] of the different ventilation terms, for today (a) and three hypothesized ice-age conditions ((b), (c),
and (d)); calculation of the mean preformed [PO32
4 ] of the ocean interior for each of these cases (e); and the effects of
the hypothesized glacial changes (b), (c), and (d) on atmospheric CO2 compared with the change in mean preformed
[PO32
4 ], as calculated by the CYCLOPS ocean geochemistry box model (f) (Keir, 1988; Sigman et al., 1998). In the
interglacial case (a), there are three sources of subsurface water shown: North Atlantic Deep Water (NADW),
Antarctic Intermediate Water (AAIW), and Antarctic Bottom Water (AABW). In this simple picture, NADW and
AAIW formation are related; NADW is drawn into the Antarctic surface by the wind-driven (“Ekman”) divergence of
surface waters and then forms AAIW because of surface-water convergence further north (Gordon et al., 1977;
Toggweiler and Samuels, 1995). The rates of formation (in Sverdrups, 106 m3 s21) and preformed [PO32
4 ] (in mM) are
from the CYCLOPS model. In (b), stratification is assumed to shut off AABW formation. Because we do not change
NADW, we are forced to leave AAIW unchanged as well. In (b), export production in the Antarctic decreases
proportionally with the decrease in upwelling to the surface, such that surface [PO32
4 ] remains at 1.7 mM. This would
be the expected outcome if, for instance, the Antarctic is iron limited and most of the iron comes from upwelling,
during both interglacials and ice ages. In (d), the same circulation change occurs as in (b), but in this case export
production does not decrease as upwelling decreases, so that surface [PO32
4 ] drops to ,1.3 mM. This would be the
expected outcome if, for instance, the Antarctic iron is supplemented by a high dust input during ice ages. In (c),
circulation remains constant but Antarctic export production increases (as indicated by the increase in the downward
green arrow). This would be the expected outcome if, for instance, the combined input of iron from dust supply and
deep water is greater during ice ages (Martin, 1990; Watson et al., 2000). In (e), the mean preformed [PO32
4 ] of the
ocean interior is calculated as the total phosphate input to the ocean interior divided by the ventilation rate (North
Atlantic terms on the left, Antarctic terms on the right). In (f), the calculated change in atmospheric pCO2 (relative to
the interglacial case (a)) is plotted against the mean preformed [PO32
4 ] of each case, as calculated in (e). The strong
correlation between these two parameters (a 170 ppm decrease in CO2 for a 1 mM decrease in [PO32
4 ]) indicates that
the mean preformed [PO32
4 ] is an excellent predictor for the effect of a given change on the strength of the biological
pump. In these experiments, the mean gas exchange temperature for the ocean interior was held constant, so that no
change in the “solubility pump” was allowed (see text). In addition, calcium carbonate dynamics are included in these
model experiments, so that a fraction of the CO2 response (,25%) is not the result of the biological pump, in the most
strict sense. Finally, none of these experiments address the evidence that NADW formation was reduced during
glacial times (Boyle and Keigwin, 1982).
preformed nutrients (Figures 8(b) and (d)).
Increasing surface productivity would lower
the preformed nutrient concentration of the deep
sea solely by reducing the preformed nutrient
concentration of the Southern Ocean contribution
to deep water (Figure 8(c)). Stratification of the
Southern Ocean surface would reduce the fraction
of the deep ocean ventilated by the Southern
Ocean, allowing the ocean interior to migrate
toward the lower preformed nutrient concentration
508
The Biological Pump in the Past
Figure 8 (continued).
of northern-sourced deep water (Figure 8(b)).
In addition, if upon stratification phytoplankton
growth rate does not decrease as much as does the
supply of nutrients to the surface, then stratification would also lower the preformed nutrient
concentration of southern-sourced deep waters
(Figure 8(d)). In this case, stratification lowers the
preformed nutrient concentration by both (i) and
(ii) above, leading to a lower net preformed
nutrient concentration for the ocean interior and
thus making it a more potent mechanism for
lowering atmospheric CO2. Of course, there are
important additional considerations. First, a
decrease in Southern Ocean deep water might
allow the deep-sea temperature to rise, which
would push CO2 back into the atmosphere. This
apparently did not occur during glacial times
(Schrag et al., 1996), suggesting that all glacialage deep-water sources were very cold. Second,
and more importantly, North Atlantic deep water
formation may have been weaker during the last
ice age (Boyle and Keigwin, 1982) (see Chapter
6.16). As this deep-water source is the low
preformed nutrient end-member today, its
reduction would have worked to increase the
preformed nutrient concentration of the deep
ocean during the last ice age, potentially offsetting
the effects of a decrease in Southern-sourced deep
water. These questions about actual events aside,
the mean preformed nutrient concentration of the
deep ocean provides an important predictive index
for the effect of changes in the polar ocean on
atmospheric CO2.
Some attention has focused on the prevention
of CO2 release from the Antarctic surface by sea
ice cover as a potential driver of glacial/interglacial CO2 change (Stephens and Keeling, 2000).
Though the mechanism is essentially physical, it
nevertheless would cause an increase in the global
efficiency of the biological pump by removing
its polar leak. This is an important reminder that
the biological pump is not solely a biological
process, but rather arises from the interaction
between the biology and physics of the ocean. For
prevention of CO2 release by sea ice cover to be
the sole mechanism for glacial/interglacial CO2
change, nearly complete year-round ice coverage
of the Antarctic would be required. This seems
unlikely to most investigators (Crosta et al.,
1998). As a result, a hybrid hypothesis has been
proposed for the glacial Antarctic, in which
intense surface stratification and nutrient consumption during the summer was followed by
wintertime prevention of gas exchange due to ice
cover (Moore et al., 2000). While promising, this
mechanism faces a discrepancy with the evidence
for lower export production in the glacial
Antarctic (see below). In the modern Antarctic,
nutrients are supplied to the surface largely by
wintertime vertical mixing. Thus, without yearround stratification, higher annual export production would have been required to lower the
surface nutrient concentration below current
summertime levels, even for a brief summer
period. An alternative combined gas exchange
limitation/stratification mechanism that does not
violate the productivity constraint is that of
spatial segregation of the two mechanisms within
the Antarctic (Figure 9). Near the Antarctic
continental margin, the formation of thick ice
cover and its associated salt rejection could drive
overturning, but the near-complete ice cover in
that region could prevent CO2 release. The ice
formed along the coast would eventually drift into
the open Antarctic, where it would melt, thus
stratifying the water column and preventing the
release of CO2 from open Antarctic as well. In
any case, ice cover provides an alternative or
Tools
509
6.18.3 TOOLS
Figure 9 North-South (left-to-right) depth section of
the shallow Antarctic, (a) under modern conditions, and
(b) under glacial conditions as posited by a hybrid
hypothesis involving both limitation of gas exchange by
sea ice near the Antarctic continent and sea ice meltdriven stratification of the open Antarctic. During
glacial times, sea ice formation was likely vigorous in
the Antarctic, and the salt rejection during freezing
would have worked to increase the density of the
surface, perhaps driving vigorous overturning (bold
loop close to the Antarctic continent) (Keeling and
Stephens, 2001). While this alone would act to ventilate
the deep sea and release CO2 to the atmosphere, the ice
coverage may have been adequately dense to prevent
CO2 outgassing, which occurs more slowly than the
exchange of heat and O2 (Stephens and Keeling, 2000).
Here, we posit that this region of ice formation and
ventilation was limited to the waters close to the
Antarctic continent. The winds in the Antarctic would
have drawn the ice northward, and some of it would
have melted in the open Antarctic. This would have
worked to strengthen the vertical stratification in the
open Antarctic, which may have reduced CO2 loss from
the ocean in this region (Francois et al., 1997).
Depending on the predominant route of deep-ocean
ventilation in the modern Antarctic (coastal or open
ocean, see text), either the coastal sea ice coverage or
the open Antarctic stratification may be taken as the
most important limitation on CO2 release. This hybrid
hypothesis has a significant list of requirements in order
to be feasible. For instance, the surface waters in the
hypothesized region of convection must not mix
laterally with the surface waters of the open Antarctic.
additional potential mechanism for preventing the
release of biologically sequestered CO2 from the
Antarctic, and future paleoceanographic work
must try to provide constraints on its actual
importance.
While there are many aspects of biological
productivity that one might hope to reconstruct
through time, there are two parameters that are most
fundamental to the biological pump: (i) export
production and (ii) nutrient status. In regions where
a deviation from major-nutrient limitation is highly
unlikely, in particular, the tropical and subtropical
ocean, export production is the major regional
constraint that local paleoceanographic data can
provide. For instance, if export production in the
tropics was higher some time in the past, then there
are at least two plausible explanations: (i) there was
more rapid water exchange between the surface and
the nutrient-bearing subsurface water, or (ii) there
was an increase in the nutrient content of the
subsurface water (and/or in N2 fixation in the
surface; see discussion above). While physically
driven changes in nutrient supply (e.g., changes in
upwelling intensity) have limited significance for
the biological pump, changes in deep-ocean nutrient
content could drive a strengthening of the pump and
thus an associated decline in atmospheric CO2 (see
discussion above). Thus, it is critical for studies of
the biological pump not only to reconstruct changes
in low-latitude export production but also to
determine what caused the changes.
In the polar ocean, both export production and
nutrient status must be known to determine the
impact on atmospheric CO2, because both of these
terms are needed to determine the ratio of CO2
supply from deep water to CO2 sequestration by
export production (Figure 7). For instance, a
decrease in export production associated with an
increase in surface nutrients would imply an
increased leak in the biological pump, whereas a
decrease in export production associated with
reduced nutrient availability would imply a
smaller leak in the pump. In low-latitude regions
of upwelling, nutrient status is less important from
the perspective of the global biological pump.
Nevertheless, since nutrient status is potentially
variable in these environments, it must be
constrained to develop a tractable list of explanations for an observed change in productivity.
A great many approaches have been taken to
study the biological pump in the past. While there
are approaches that take advantage of virtually
every aspect of the available geologic archives, we
focus here on tools to reconstruct export production
and nutrient status and limit ourselves to those that
are largely based on geochemical measurement. We
refer the interested reader to Fischer and Wefer
(1999) for additional information and references.
6.18.3.1 Export Production
Export production, as defined above, is the flux
of organic carbon that sinks (or is mixed) out of
510
The Biological Pump in the Past
the surface ocean and into the deep sea. While it is
extremely difficult to imagine how paleoceanographic measurements can provide direct
constraints on this highly specific parameter,
approaches exist for the reconstruction of the
flux of biogenic debris to the seafloor. To the
degree that sinking biogenic debris has a predictable chemical composition and that its export
out of the surface ocean is correlated with its rain
rate to the seafloor, these approaches can provide
insight into export production variations (Muller
and Suess, 1979; Ruhlemann et al., 1999;
Sarnthein et al., 1988) (see Chapter 6.04).
Only a very small fraction of the organic carbon
exported out of the surface ocean accumulates in
deep-sea sediments, with loss occurring both in the
water column and at the seafloor. Moreover,
various environmental conditions may influence
the fraction of export production preserved in the
sedimentary record (see Chapter 6.11). The
mineral components of the biogenic rain (calcium
carbonate, opal) do not represent a source of
chemical energy to the benthos and thus may be
preserved in a more predictable fashion in deepsea sediments. Thus, paleoceanographers sometimes hope to reconstruct export production (the
flux of organic carbon out of the surface ocean)
from the biogenic rain of opal or carbonate to the
seafloor (Ruhlemann et al., 1999). We know that
systematic variations in the ratio of opal or calcium
carbonate to organic carbon in export production
do occur, so there are many cases where changes
in the mineral flux can be interpreted either as a
change in the rate of export production or in its
composition. Despite these limitations, we are
better off with this information than without it.
The most basic strategy to reconstruct the
biogenic rain to the seafloor is to measure
accumulation rate in the sediments. This approach
is appropriate for materials that accumulate without loss in the sediments or that are preserved to a
constant or predictable degree. For the biogenic
components of interest, organic carbon, opal, and
calcium carbonate, the degree of preservation
varies with diverse environmental variables,
detracting from the usefulness of accumulation
rate for the reconstruction of their rain to the
seafloor. Nevertheless, because the environmental
variables controlling preservation in the sediments
are often linked to rain rate to the seafloor,
accumulation rate can sometimes be a sound basis
for at least the qualitative reconstruction of
changes in some components of the biogenic
rain. For instance, the fraction of the opal rain that
is preserved in the sediment appears to be higher
in opal-rich environments (Broecker and Peng,
1982). As a result, an observed increase in
sedimentary opal accumulation rate may have
been partially due to an increase in preservation.
However, an increase in opal rain (and thus
diatom-driven export) would have been needed to
increase the sedimentary opal content in the first
place.
If an age model can be developed for a deep-sea
sediment core, then the accumulation histories of
the various sediment components can be reconstructed. Assuming that the age model is correct,
the main uncertainty in the environmental significance of the reconstruction is then the potential for
lateral sediment transport. Sediments can be
winnowed from or focused to a given site on the
seafloor.
The geochemistry of radiogenic thorium provides a way to evaluate the effect of lateral
sediment transport and related processes (Bacon,
1984; Suman and Bacon, 1989) (see Chapter
6.09). Thorium-230 (230Th) is produced at a
constant rate throughout the oceanic water column
by the decay of dissolved 234U. As 230Th is
produced, it is almost completely scavenged onto
particles at the site of its production. As a result,
the accumulation rate of 230Th in deep-sea
sediments should match its integrated production
in the overlying water column. If the accumulation
of 230Th over a time period, as defined by a
sediment age model, is more or less than should
have been produced in the overlying water column
during that period, then sediment is being focused
or winnowed, respectively.
Thorium-230 is also of great use as an
independent constraint on the flux of biogenic
material to the seafloor (Bacon, 1984; Suman and
Bacon, 1989). Because the production rate of
230
Th in the water column is essentially constant
over space and time, its concentration in sinking
particles is diluted in high-flux environments with
a large biogenic flux to the seafloor, yielding
lower sedimentary concentrations of 230Th in
these environments. One limitation on the use of
230
Th is the radioactive decay of this isotope,
which has a half-life of 7.5 £ 104 yr. Helium-3
can apparently be used in similar ways to 230Th
and has the advantage that it is a stable isotope
(Marcantonio et al., 1996); however, it has been
studied less than 230Th.
The addition of other scavenged elements further
enhances the utility of 230Th. 231Pa, like 230Th, is
produced throughout the ocean water column by
radioactive decay of a uranium isotope (235U in the
case of 231Pa). However, protactinium is somewhat
less particle-reactive than thorium. As a result, the
downward flux of 231Pa to the seafloor varies across
the ocean. Protactinium from low-flux regions
mixes into high-flux regions to be scavenged,
resulting in higher 231Pa fluxes in environments
with high biogenic rain rates. Thus, the 231Pa/230Th
ratio of sediments has been studied as an index of the
particle flux through the water column. This index is
potentially complementary to thorium-normalized
accumulation rates in that the 231Pa/230Th ratio
Tools
should not be affected by losses during sedimentary
diagenesis. Limitations of this approach include
(i) the tendency for the less easily scavenged isotope
(231Pa) to adsorb preferentially to certain types of
particles (e.g., diatom opal), which can make it
difficult to distinguish a change in the flux of
particles from a compositional change, and (ii) the
scavenging of protactinium on basin margins, which
can vary in importance as ocean circulation changes
(Chase et al., 2003b; Walter et al., 1999; Yu et al.,
2001). Beryllium-10 may be used in similar ways to
231
Pa but has been studied less (Anderson et al.,
1990).
The flux of barium to the seafloor appears to be
strongly related to the rain of organic matter out of
the surface ocean (Dehairs et al., 1991). Apparently, the oxidation of organic sulfur produces
microsites within sinking particles that become
supersaturated with respect to the mineral barite
(BaSO4). On this basis, barium accumulation has
been investigated and applied as a measure of
export production in the past, representing a more
durable sedimentary signal of the organic carbon
sinking flux than sedimentary organic carbon
itself (Dymond et al., 1992; Gingele et al., 1999;
Paytan et al., 1996). While debates continue on
aspects of the biogenic barium flux, some problems
with preservation are broadly recognized. In
sedimentary environments with low bottom water
O2 and/or high organic carbon rain rates, active
sulfate reduction in the shallow sediments can
cause the barite flux to dissolve. In addition, there is
some low level of barite dissolution under all
conditions; if the biogenic barium flux is low, a
large fraction of the barite can dissolve at the
seafloor. Thus, barium accumulation studies
appear to be most applicable to environments of
intermediate productivity.
The rapidly growing field of organic geochemistry promises new approaches for the study of
biological productivity in the past. By studying
specific chemical components of the organic matter
found in marine sediments, uncertainties associated
with carbon source can be removed, and a richer
understanding of past surface conditions can be
developed (e.g., Hinrichs et al., 1999; Martinez
et al., 1996). To some degree, organic geochemical
approaches may allow us to circumvent the thorny
problem of breakdown and alteration at the seafloor
(Sachs and Repeta, 1999).
6.18.3.2 Nutrient Status
The measurable geochemical parameters currently available for addressing the nutrient status
of the surface ocean include (i) the Cd/Ca
ratio (Boyle, 1988a) and 13C/12C of planktonic
(surface-dwelling) foraminiferal carbonate
(Shackleton et al., 1983), which are intended to
record the concentration of cadmium and the
13
511
12
C/ C of DIC in surface water, (ii) the nitrogen
isotopic composition of bulk sedimentary organic
matter (Altabet and Francois, 1994a) and microfossil-bound organic matter (Shemesh et al., 1993;
Sigman et al., 1999b), which may record the
degree of nitrate utilization by phytoplankton in
surface water, (iii) the silicon isotopic composition (De La Rocha et al., 1998) and Ge/Si ratio
(Froelich and Andreae, 1981) of diatom microfossils, which may record the degree of silicate
utilization in surface water (but see Bareille et al.
(1998)), and (iv) the carbon isotopic composition
of organic matter in sediments (Rau et al., 1989)
and diatom microfossils (Rosenthal et al., 2000;
Shemesh et al., 1993), which may record the
dissolved CO2 concentration of surface water and/
or the carbon uptake rate by phytoplankton.
The concentration of dissolved cadmium is
strongly correlated with the major nutrients
throughout the global deep ocean, and the Cd/Ca
ratio of benthic (seafloor-swelling) foraminifera
shells records the cadmium concentration of
seawater (Boyle, 1988a). As a result, benthic
foraminiferal Cd/Ca measurements allow for the
reconstruction of deep-ocean nutrient concentration gradients over glacial/interglacial cycles.
Planktonic foraminiferal Cd/Ca measurements in
surface water provide an analogous approach to
reconstruct surface ocean nutrient concentration,
although this application has been less intensively
used and studied.
A number of factors appear to complicate the
link between planktonic foraminiferal Cd/Ca and
surface nutrient concentrations in specific regions.
First, temperature may have a major effect on the
Cd/Ca ratio of planktonic foraminifera (Elderfield
and Rickaby, 2000; Rickaby and Elderfield,
1999). Second, planktonic foraminiferal shell
growth continues below the surface layer (Bauch
et al., 1997; Kohfeld et al., 1996) and probably
integrates the Cd/Ca ratio of surface and shallow
subsurface waters. This is of greatest concern in
polar regions such as the Antarctic, where there
are very sharp vertical gradients within the depth
zone in which planktonic foraminferal calcification occurs. Third, carbonate geochemistry on the
deep seafloor may affect the Cd/Ca ratio of
foraminifera preserved in deep-sea sediments
(McCorkle et al., 1995). Finally, the link between
cadmium and phosphate concentration in surface
waters is not as tight as it is in deeper waters (Frew
and Hunter, 1992).
Because of isotopic fractionation during carbon
uptake by phytoplankton, there is a strong
correlation between the 13C/12C of DIC and
nutrient concentration in deep waters; as a result,
the 13C/12C of benthic foraminiferal fossils is a
central tool in paleoceanography. The 13C/12C of
planktonic foraminiferal calcite has been used
as a tool to study the strength of the biological
512
The Biological Pump in the Past
pump (Shackleton et al., 1983). However, the
exchange of CO2 with the atmosphere leads to a
complicated relationship between the 13C/12C of
DIC and the nutrient concentration in surface
waters (Broecker and Maier-Reimer, 1992), such
that even a perfect reconstruction of surface water
13 12
C/ C would not provide direct information of
surface-ocean nutrient status. In addition, the
13 12
C/ C of planktonic foraminiferal fossils found
in surface sediments appears to be an imperfect
recorder of the 13C/12C of DIC in modern surface
waters, for a variety of reasons (Spero et al., 1997;
Spero and Lea, 1993; Spero and Williams, 1988).
Finally, the same concerns about the calcification
depth noted for Cd/Ca also apply to carbon
isotopes or, for that matter, any geochemical
signal in planktonic foraminiferal calcite.
During nitrate assimilation, phytoplankton
preferentially consume 14N-nitrate relative to
15 N-nitrate (Montoya and McCarthy, 1995;
Waser et al., 1998), leaving the surface nitrate
pool enriched in 15N (Sigman et al., 1999a). This
results in a correlation between the 15N/14N ratio
of organic nitrogen and the degree of nitrate
utilization by phytoplankton in surface waters
(Altabet and Francois, 1994a,b). There are major
uncertainties in the use of this correlation as the
basis for paleoceanographic reconstruction of
nutrient status, which include (i) the isotopic
composition of deep-ocean nitrate through time,
(ii) temporal variations in the relationship between
nitrate utilization and the nitrogen isotopes in
the surface ocean (i.e., the “isotope effect” of
nitrate assimilation), and (iii) the survival of the
isotope signal of sinking organic matter into the
sedimentary record (e.g., Lourey et al., in press).
The nitrogen isotope analysis of microfossilbound organic matter (Sigman et al., 1999b) and
of specific compound classes such as chlorophylldegradation products (Sachs and Repeta, 1999)
promises to provide tools to evaluate and
circumvent the effect of diagenesis. However, it
remains to be seen whether selective nitrogen
pools such as microfossil-bound nitrogen are
tightly linked to the nitrogen isotope ratio of the
integrated sinking flux, the parameter that theoretically relates most directly to the degree of
nitrate utilization in surface waters (Altabet and
Francois, 1994a).
The isotopic composition of silicon in diatom
opal has been investigated as a proxy for the
degree of silicate utilization by diatoms, based on
the fact that diatoms fractionate the silicon
isotopes (30Si and 28Si) during uptake (De La
Rocha et al., 1997, 1998). This application is
analogous to the use of nitrogen isotopes to study
nitrate utilization, but with important differences.
On the one hand, the upper ocean cycle is arguably
simpler for silica than bio-available nitrogen,
which bodes well for the silicon isotope system.
On the other hand, there are very few regions of
the surface ocean that maintain high dissolved
silicate concentrations (Figure 4(d)). As a result,
in regions of strong silicate gradients, the link
between silicate utilization and silicon isotopic
composition may be compromised by mixing
processes in surface waters.
The carbon isotopic composition of sedimentary organic matter was originally developed as a
paleoceanographic proxy for the aqueous CO2
concentration of Southern Ocean surface waters
(Rau et al., 1989). The aqueous CO2 concentration
is a nearly ideal constraint for understanding a
region’s effect on the biological pump, as it would
provide an indication of its tendency to release or
absorb carbon dioxide. However, it has become
clear that growth rate and related parameters are
as important as the concentration of aqueous CO2
for setting the 13C/12C of phytoplankton biomass
and the sinking organic matter that it yields (Popp
et al., 1998). In addition, active carbon acquisition
by phytoplankton is also probably important for
phytoplankton 13C/12C in at least some environments (Keller and Morel, 1999). Thus, the 13C/12C
of organic carbon is a useful paleoceanographic
constraint (Rosenthal et al., 2000), but one that is
currently difficult to interpret in isolation.
6.18.3.3 Integrative Constraints on
the Biological Pump
Above, we have focused on approaches to
reconstruct the role of a specific region of the
surface ocean on the global biological pump.
However, if our goal is to explain the global
net effect of ocean biology on the carbon cycle,
we must also search for less local, more
integrative constraints on the biological pump.
This is possible because the atmosphere, surface
ocean and deep sea are each relatively homogenous geochemical reservoirs, while being
distinct from one another. There are a number
of global scale geochemical parameters that
may provide important constraints on the
biological pump; we describe several of these
below.
6.18.3.3.1
Carbon isotope distribution of
the ocean and atmosphere
As described above, the biological pump tends
to sequester 12C-rich carbon in the ocean interior.
All else being equal, the stronger the global
biological pump, the higher will be the 13C/12C of
dissolved inorganic carbon in the surface ocean
and of carbon dioxide in the atmosphere. Broecker
(1982a,b) and Shackleton (see Shackleton et al.,
1983) compared sediment core records of
Tools
13
12
the C/ C of calcite precipitated by planktonic
and benthic foraminfera, the goal being to
reconstruct the 13C/12C difference in DIC between
the surface ocean and the deep sea, a measure of
the strength of the global ocean’s biological pump.
Indeed, this work was the first suggestion that the
biological pump was stronger during ice ages, thus
potentially explaining the lower CO2 levels of
glacial times. Our view of these results is now
more complicated (e.g., Spero et al., 1997);
however, the basic conclusion remains defensible
(Hofmann et al., 1999). Reliable measurement of
the 13C/12C of atmospheric CO2 has proven
challenging (Leuenberger et al., 1992; Marino
and McElroy, 1991; Smith et al., 1999). Moreover, there are additional modifiers of the 13C/12C
of atmospheric CO2, such as the temperature of
gas exchange. Nevertheless, these data also seem
consistent with the biological pump hypothesis for
glacial/interglacial CO2 change (Smith et al.,
1999).
6.18.3.3.2
Deep-ocean oxygen content
The atmosphere/ocean partitioning of diatomic
oxygen (O2) is a potentially important constraint
on the strength of the biological pump; the
stronger the pump, the more O2 will be shuttled
from the deep ocean to the surface ocean and
atmosphere. The rain of organic detritus into the
deep ocean drives an O2 demand by the deepocean benthos as it sequesters carbon dioxide in
the deep sea, while exposure of deep waters at the
surface recharges them with O2 as it allows deep
waters to degas excess CO2 to the atmosphere
(Figure 6). A decrease in atmospheric CO2 due to
the biological pump should, therefore, be
accompanied by a decrease in the O2 content of
the ocean subsurface.
The concentration of dissolved O2 in the ocean
interior has long been a target for paleoceanographic reconstruction (the atmospheric change in
O2 content would be minute). Sediments underlying waters with nearly no O2 tend to lack
burrowing organisms, so that sediments in these
regions are undisturbed by “bioturbation” and can
be laminated; this is perhaps our most reliable
paleoceanographic indicator of deep-water O2
content. Arguments have been made for surface
area-normalized sedimentary organic carbon content as an index of O2 content in some settings
(Keil and Cowie, 1999). It remains to be seen
whether this is complicated by the potential for
changes in the rain rate of organic matter to the
sediments. There are a number of redox-sensitive
metals, the accumulation of which gives information on the O2 content of the pore water in
shallow sediments (Anderson et al., 1989; Crusius
et al., 1996; Crusius and Thomson, 2000).
Unfortunately, the O2 content of the sediment
513
pore waters can vary due to organic matter supply
to the sediments as well as the O2 content of the
bottom water bathing the seafloor, so that these
two parameters can be difficult to separate (a
situation that is analogous to that for sedimentary
organic carbon content). While interesting data
and arguments have been put forward in support
of various approaches (Hastings et al., 1996;
Russell et al., 1996), it seems fair to argue that the
paleoceanographic community still lacks a
reliable set of methods for the global reconstruction of deep ocean dissolved O2 content,
especially in environments far from complete O2
consumption.
Initial model results suggested that a biological
pump mechanism for the glacial/interglacial CO2
change would have rendered the ocean subsurface
so O2 deficient as to prevent the presence of
burrowing organisms and oxic respiration over
large expanses of the seafloor, which should leave
some tell-tale signs in the sediment record.
However, observations have changed this story
significantly. For instance, the O2 minimum,
which is at intermediate depths in the modern
ocean, may have migrated downward into the
abyssal ocean during the last ice age (Berger and
Lange, 1998; Boyle, 1988c; Herguera et al., 1992;
Marchitto et al., 1998), so that the O2 decrease
was apparently focused in waters which are
currently relatively rich in O2, perhaps avoiding
widespread anoxia at any given depth (Boyle,
1988c). For this reason, testing the biological
pump hypothesis by reconstructing deep ocean O2
will require that we do more than simply search
for extensive deep-sea anoxia; rather, it will
probably require a somewhat quantitative indicator of dissolved O2 that works at intermediate
O2 concentrations.
6.18.3.3.3
Phasing
With adequate dating and temporal resolution
in paleoceanographic and paleoclimatic records,
the sequence of past events and changes can be
reconstructed, providing among the most direct
evidence for cause and effect. There is much
information on phasing that is relevant to the
biological pump and its role in glacial/interglacial
CO2 change. We limit ourselves here to one
example that arises largely from ice core records:
the timing of CO2 change relative to temperature
and glaciation (Broecker and Henderson, 1998
and references therein). Near the end of ice ages,
atmospheric CO2 rises well before most of the
deglaciation (Monnin et al., 2001; Sowers and
Bender, 1995); this was referred to above as
strong evidence against Broecker’s shelf phosphorus hypothesis for CO2 change (see Section
6.18.2.1). While the phasing of temperature is
514
The Biological Pump in the Past
still debated, it appears that most of the warming
in the high-latitude southern hemisphere preceded
most of the warming in the high-latitude northern
hemisphere, and that the atmospheric CO2 rise
lagged only slightly behind the southern hemisphere warming. This latter observation is roughly
consistent with a variety of hypotheses for
changes in the biological pump, but appears
inconsistent with alternative hypotheses that
rely solely on changes in the oceanic calcium
carbonate budget, which operates on a longer
timescale than the biological pump (Archer and
Maier-Reimer, 1994; Opdyke and Walker, 1992).
With continued work, the detailed timing of CO2
change may provide quantitative constraints on
changes in the oceanic calcium carbonate budget
as a partial contributor to CO2 change; calcium
carbonate plays a role in many of the biological
pump hypotheses (Archer et al., 2000a; Sigman
et al., 1998; Toggweiler, 1999).
6.18.4 OBSERVATIONS
To this point, a number of central observations
and concepts have already been described to
motivate a search for changes in the biological
pump over glacial/interglacial cycles. First, there
are carbon dioxide variations over glacial/interglacial cycles that are of the right magnitude and
temporal structure to be caused by changes in the
biological pump. Second, carbon isotope data for
carbon dioxide and foraminiferal carbonate
appear consistent with a biological-pump mechanism. Third, changes in the nitrogen cycle have
been recognized that would tend to increase the
oceanic nitrate reservoir during glacial times,
although an actual increase in this reservoir has
not been demonstrated; such a reservoir change
might be expected to strengthen the low-latitude
pump during glacial times. Finally, observations
about phytoplankton in polar regions, in particular, their incomplete consumption of the major
nutrients and their tendency toward iron limitation, suggest that simple changes in either the
iron supply to the Antarctic surface ocean or in
polar ocean circulation could lead to an increase in
the efficiency of the high-latitude biological pump
during glacial times. These observations, together
with other ideas about the operation of the ocean
carbon cycle, warrant that we consider the basic
regional observations on biological productivity
and nutrient status over glacial/interglacial cycles.
6.18.4.1 Low- and Mid-latitude Ocean
There have been many studies of the response
of coastal and equatorial upwelling systems
to glacial/interglacial climate change, and this
overview cannot do justice to the entire body of
work. The coastal upwelling regions along the
western continental margins show an Atlantic/
Pacific difference in their response to glacial/
interglacial climate change. Studies along the
western coast of Africa suggest an increase in
productivity during glacial times (Summerhayes
et al., 1995; Wefer et al., 1996). However, in the
eastern Pacific, the coastal upwelling zones off of
California and Mexico in the north and off of
Peru in the south were less productive during the
last glacial period (Dean et al., 1997; Ganeshram
and Pedersen, 1998; Ganeshram et al., 2000;
Heinze and Wefer, 1992). This response has
generally been explained as the effect of continental cooling (and a large North American ice
sheet in the case of the California Current) on the
winds that currently drive coastal upwelling in the
eastern tropical Pacific (Ganeshram and Pedersen,
1998; Herbert et al., 2001).
Early paleoceanographic studies in the equatorial Pacific found that export production was
greater in the equatorial Pacific during ice ages
(Lyle, 1988; Pedersen, 1983; Pedersen et al.,
1991; Rea et al., 1991), and some subsequent
studies have supported this conclusion (Herguera
and Berger, 1991; Murray et al., 1993; Paytan
et al., 1996; Perks et al., 2000). However, this
interpretation has been put in question
by reconstructions using 230Th- and 3Henormalized accumulation rates, 231Pa/230Th ratios
(Marcantonio et al., 2001b; Schwarz et al., 1996;
Stephens and Kadko et al., 1997) and other
evidence (Loubere, 1999). From studies along
the eastern equatorial Atlantic, the consensus is
for higher productivity during the last ice age than
the Holocene; because of the limited extent of the
Atlantic basin, it is difficult to differentiate this
change from the glacial-age increase in productivity in African coastal upwelling (Lyle
et al., 1988; Martinez et al., 1996; Moreno et al.,
2002; Rutsch et al., 1995; Schneider et al., 1996).
Proxies for nutrient utilization (Altabet, 2001;
Farrell et al., 1995; Holmes et al., 1997), pH
(Sanyal et al., 1997), and wind strength (Stutt
et al., 2002) would suggest that any export
production increases that did occur in the lowlatitude upwelling systems of the Pacific and
Atlantic during the last glacial maximum were due
to higher wind-driven upwelling, which increased
the nutrient supply to the surface.
Upwelling associated with monsoonal circulation has apparently behaved predictably since the
last ice age (Duplessy, 1982; Prell et al., 1980).
With less summertime warming of the air over
Asia during the last glacial maximum, there was a
weakening of the southwesterly winds of the
summertime monsoon that drive coastal upwelling
off the horn of Africa and Saudi Arabia, leading to
a glacial decrease in export production in that
Observations
region (Altabet et al., 1995; Anderson and Prell,
1993; Marcantonio et al., 2001a; Prell et al.,
1980). To the east and in the more open Indian
Ocean, where the (southward) wintertime monsoonal winds drive upwelling or vertical mixing,
productivity was apparently higher during the last
glacial maximum, again consistent with the
expected effect of a cooler Eurasian climate on
the monsoon cycle (Beaufort, 2000; Duplessy,
1982; Fontugne and Duplessy, 1986). In the South
China Sea, where the winter monsoon (seaward
flow) is responsible for the exposure of nutrientrich deep waters, export production was apparently higher during the last glacial maximum,
again suggesting a stronger winter monsoon
associated with colder winter conditions over
Eurasia (Huang et al., 1997; Thunell et al., 1992).
The effect of the wintertime monsoon winds may
also have impacted the western equatorial Pacific
(Herguera, 1992; Kawahata et al., 1998).
The low-productivity regions of the low- and
mid-latitude ocean are an important source of
information on the strength of the low-latitude
biological pump, as they may be less susceptible
to wind-driven changes in nutrient supply than are
the upwelling regions discussed above, so that
changes in productivity would be more closely
tied to changes in the oceanic nutrient reservoir
(and/or in N2 fixation). As mentioned above,
studies in the western equatorial Pacific indicate
that the sedimentary Corg was higher in these
regions during glacial times (Figure 10; Perks
et al. (2000)); however, studies from the more
southern western Pacific appear to suggest lower
productivity (Kawahata et al., 1999). Studies in
the western equatorial and tropical Atlantic
indicate lower CaCO3 and Corg burial rates during
glacial times (François et al., 1990; Ruhlemann
et al., 1996). In the tropical Indian Ocean, away
from the Arabian Sea, productivity was apparently
higher during glacial times, but this has again been
explained as the result stronger wind-driven
mixing during the wintertime monsoon (Beaufort,
2000; Fontugne and Duplessy, 1986). Discrete
sedimentation events of massive mat-forming
diatoms are found in glacial-age sediments from
the equatorial Indian; their significance is unclear
(Broecker et al., 2000).
As described in Section 6.18.2, there is
substantial evidence for a change in the spatial
pattern of denitrification across glacial/interglacial
transitions (Altabet et al., 1995, 2002; Emmer and
Thunell, 2000; Ganeshram et al., 1995; Pride et al.,
1999), and it seems likely that this led to a net
global decrease in the loss rate of nitrate from the
ocean during the last glacial maximum. Depending on the changes in N2 fixation over glacial/
interglacial cycles and the role of phosphorus in
these changes (see Section 6.18.2), low-latitude
productivity may or may not have responded so as
515
to strengthen the biological pump. As discussed
above, increased low-latitude productivity due to
increased upwelling or vertical mixing has limited
significance for the strength of the biological
pump. It is a challenge to develop an approach to
study past productivity changes that can distinguish between a change in the physical rate of
vertical exchange and a change in the nutrient
content of the subsurface. Thus, it is difficult to
test the hypothesis that the ocean nutrient
reservoir drove a significant glacial increase in
the low- and mid-latitude biological pump, so as
to explain glacial/interglacial CO 2 changes.
Nevertheless, it is notable that no direct support
for the hypothesis has arisen without a focused
search. Based on the overview above, the
evidence for a net global increase in low-latitude
productivity during the last ice age is not
compelling. Moreover, the evidence that does
exist for regional increases in low-latitude productivity seems easily explained in terms of
changes in the wind-driven upwelling. Comparison of accumulation records for organic carbon
and opal (Herguera and Berger, 1994) and of
records from opposite sides of the South Atlantic
basin (Ruhlemann et al., 1999) has led to the
suggestion that the nutrient content of the
thermocline was lower during ice ages. If such
was indeed the case, it could be associated with a
tendency for nutrients to shift into the deeper
ocean during ice ages (Boyle, 1988b,c); nevertheless, it provides little encouragement for
hypotheses of a larger oceanic nutrient reservoir
during glacial times.
While it is tempting to phrase orbitally driven
climate variability in terms of “ice ages” and
“interglacials,” this simplification breaks down
frequently in the lower-latitude ocean (Clemens
et al., 1991; McIntyre and Molfino, 1996; Sachs
et al., 2001). The precession component of the
Earth’s orbital variations is by far the most
directly important for the energy budget of the
tropics. The Pleistocene variability of productivity
in some tropical regions (Pailler et al., 2002) and
coastal and equatorial upwellings (Moreno et al.,
2002; Rutsch et al., 1995; Summerhayes et al.,
1995) is dominated by the precession (,23 kyr)
cycle, with the result that our focus on the Last
Glacial Maximum/Holocene comparison can be
misleading about the real timing of change. New
data make this case very strongly for equatorial
Pacific productivity (Perks et al., 2000). As the
variations in atmospheric CO2, glacial ice volume,
and polar temperature all have less variability at
the precession frequency, one would surmise that
the low-latitude productivity changes dominated
by precession are not major drivers of variability
in the biological pump or in atmospheric CO2.
This is consistent with the conceptual point that
wind-driven changes in low-latitude productivity
Observations
are of limited importance to CO2 transfer between
the atmosphere and ocean.
6.18.4.2 High-latitude Ocean
The global high-latitude open ocean includes the
Arctic, the North Atlantic, the Southern Ocean, and
the Subarctic North Pacific (Figure 5). We address
each of these in turn.
The modern open Arctic appears to be relatively unproductive. Annually averaged solar
insolation is low and sea ice cover is extensive,
so that light limitation of phytoplankton growth is
likely. Surface nitrate is also quite low in the
Arctic surface (Figure 5(c)), at least partially due
to a strong permanent “halocline,” or vertical
salinity gradient, that drives a strong vertical
density gradient in the upper Arctic ocean,
isolating fresh surface waters from saltier and
thus denser nutrient-rich deep water. However, as
the major nutrients are not completely absent in the
surface, the major nutrients are probably not the
dominant limitation for phytoplankton in the open
517
Arctic. The few paleoproductivity studies of the
open Arctic suggest that it was less productive
during the last ice age (Schubert and Stein, 1996;
Wollenburg et al., 2001). A range of ice age
conditions may have contributed to this decrease,
including greater ice cover and thus more extreme
light limitation as well as sea-level-driven loss of
the nutrient supply from the Pacific that currently
enters from the Bering Strait. Our current understanding of the Arctic would suggest that it is not
in itself a central part of ocean/atmosphere CO2
partitioning, so little energy has been expended to
understand its glacial/interglacial cycle.
The high-latitude North Atlantic has a productive spring bloom but tends towards majornutrient limitation in the summer. This tendency
toward nutrient limitation is due at least partially
to the formation of deep water in the high-latitude
North Atlantic. The Gulf Stream extension
provides salty, warm, low-nutrient water to the
high-latitude surface, and the cooling of this water
eventually leads to the formation of deep waters
(see Chapter 6.16). As a result of this steady state,
neither surface waters nor underlying deep waters
Figure 10 Paleoceanographic records relevant to the history of export production from various regions of the
global ocean, overlain on a map of surface nitrate concentration (in mM; Conkright et al., 1994). The records are
chosen to be somewhat representative as a crude first-order paleoproductivity record for their oceanic regime;
the extreme uncertainty associated with some of these reconstructions is discussed at length in Section 6.18.2. The
records are of the sedimentary concentration or accumulation of organic carbon or biogenic opal. While the
concentration records are complicated by potential changes in dilution by other sediment components, there are some
circumstances where accumulation rate estimates are also problematic. The referenced sources should be consulted
for information on these concerns. The biogeochemical regime of the surface ocean is roughly indicated by surface
nitrate concentration (see Figure 5), which indicates whether a record is from a nutrient-poor, low-latitude region,
a nutrient-bearing low-latitude upwelling region, or a nutrient-bearing polar region. Records are plotted versus time
(except core ODP 1094 in the Southern Ocean) and glacial –interglacial stages are numbered, with a gray shading
during interglacials. (a) The Corg accumulation record from core GeoB 1523 for the last 300 ka (Ruhlemann et al.,
1996) conforms with the general finding that productivity was lower in the oligotrophic tropical Atlantic during the
last ice age. (b) The Corg accumulation record from core GeoB 1016 for the last 300 ka (Schneider, 1991) is among the
large body of evidence indicating that the Benguela upwelling system and the eastern equatorial Atlantic were more
productive during the last ice age. However, all of the Atlantic records show a strong precessional response,
indicating only a loose connection with ice volume. Ruhlemann et al. (1999) have noted an anti-correlation between
eastern and western tropical Atlantic records. (c) Biogenic opal concentrations of ODP Site 1094 in the Antarctic zone
of Southern Ocean (Janecek, 2001) conform with the large body of data indicating lower opal accumulation rates
during ice ages. The upper 30 m of the composite sediment sequence represent approximately the last 135 ka. (d)
Biogenic opal accumulation rate in core PS2062 (Frank et al., 2000) over the last 150 ka conforms with a large body
of data indicating higher opal and organic carbon accumulation in the Subantarctic zone of the Southern Ocean during
ice ages, opposite to the temporal pattern observed in the Antarctic. (e) The Corg concentration in RNDP 74P from the
western equatorial Pacific and in ODP Site 849 from the eastern equatorial Pacific over the last 300 ka (Perks et al.,
2000) suggests higher productivity across the entire equatorial upwelling during the last ice age but shows its
dominant response to precession, not polar ice volume. (f) The Corg concentration in ODP Site 723 over the last
300 ka (Emeis et al., 1995) conforms with the large body of data suggesting lower productivity in the Oman
upwelling system during ice ages because of a weaker summer monsoon (and thus weaker upwelling) when Eurasia
is cold. (g) The Corg concentration in core NH15P during the last 120 ka in the coastal upwelling zone off
northwestern Mexico (Ganeshram and Pedersen, 1998) conforms with a large body of data indicating lower
productivity during the last ice age because the North American ice sheet caused a reduction in wind-driven coastal
upwelling. (h) Biogenic opal accumulation rates in the Subarctic North Pacific (ODP Site 882; Haug et al. (1995))
and the Okhotsk Sea (Core XP98-PC1 Narita et al. (2002)) suggest lower diatom export during glacials, perhaps
because of stronger salinity stratification of the upper water column and thus reduced nutrient supply. See text
for more references.
518
The Biological Pump in the Past
can supply large amounts of nutrients to the
surface, with the largest nutrient source being the
Antarctic Intermediate Water that penetrates from
the south. The equivocal evidence for changes in
North Atlantic productivity over glacial/interglacial cycles appears to suggest lower export
production during the last ice age (Manighetti
and McCave, 1995; Thomas et al., 1995). This is
consistent with the persistent formation of a lownutrient subsurface water in the glacial North
Atlantic (Oppo and Lehman, 1993), even though
this water mass was distinct from modern North
Atlantic Deep Water (again, see Chapter 6.16). If,
during ice ages, the North Atlantic continued to
import lower-latitude surface waters and produce
a subsurface water mass with low preformed
nutrients, then it would have played roughly the
same role in the biological pump as it does today.
However, a glacial weakening of the North
Atlantic source would have had implications for
the global mean preformed nutrient concentration
of the deep ocean (Section 6.18.2.2 and Figure 8).
Paleoceanographic work in the Southern Ocean
indicates a strong glacial/interglacial oscillation,
but with very different responses in the (poleward)
Antarctic than in the (equatorward) Subantarctic.
In the modern Antarctic, biological export production is dominated by diatoms, the group of
phytoplankton that precipitate tests of hydrated
silica (“opal”), and these microfossils are frequently the dominant component of the underlying
sediments. The accumulation of diatomaceous
sediments in the Antarctic was clearly lower
during the last ice age (Mortlock et al., 1991),
and other paleoceanographic indicators suggest
that the biological export of carbon was also lower
(Francois et al., 1997; Kumar et al., 1995;
Rosenthal et al., 2000). However, given an
apparent bias of at least some of these proxies to
preferentially record the carbon export closely
associated with diatom microfossils (Chase et al.,
2003b; Yu et al., 2001), it remains possible that the
biological export of carbon from the glacial
Antarctic surface was as high or higher than during
interglacial times but that it occurred in a form that
was poorly preserved in the sediment record
(Arrigo et al., 1999; Boyle, 1998; Brzezinski
et al., 2002; DiTullio et al., 2000; Moore et al.,
2000; Tortell et al., 2002).
In the Atlantic and Indian sectors of the
Subantarctic, there is very strong evidence for
higher export production during glacial times and
an associated increase in the relative importance of
diatom production (Chase et al., 2001; Francois
et al., 1997; Kumar et al., 1995; Mortlock et al.,
1991; Rosenthal et al., 1995). This high glacial
productivity has not yet been recognized in the
Pacific sector (Chase et al., 2003a). If this
productivity increase was indeed absent in
the Pacific sector of the Subantarctic, it provides
an important constraint on its cause. For instance,
this finding may be consistent with iron fertilization as the dominant driver of the glacial increase in
Subantarctic productivity, as the increase in iron
input may have been extremely limited in parts of
the Pacific sector (Chase et al., 2003a).
While these changes in export production are
critical to the workings of the ice age Southern
Ocean, they do not address directly the effect of
the Southern Ocean on atmospheric CO2, which is
largely determined by the competition between the
upwelling of respiratory CO2 and nutrients to
the surface and export of organic matter out of the
surface (Section 6.18.2). In the Antarctic, nitrogen
isotope data suggest that the fraction of the nitrate
consumed during the last ice age was higher than its
current value (Francois et al., 1997; Sigman et al.,
1999b). These result seem to support the longstanding hypothesis that nutrient utilization
changes in Antarctic waters are a fundamental
driver of glacial/interglacial changes in the atmospheric concentration of CO2; box model calculations predict the 25 – 40% higher nitrate
utilization (i.e., 50 – 65% in the glacial ocean
compared to 25% during the present interglacial)
could lower atmospheric CO2 by the full glacial/
interglacial amplitude under the relevant conditions (Sigman et al., 1999b). Since paleoceanographic proxy data suggest that Antarctic export
production was lower during the last ice age, one
would infer that more complete nitrate utilization
in the Antarctic was due to a lower rate of nitrate
supply from the subsurface, implying that the
fundamental driver of the CO2 change was an ice
age decrease in the ventilation of deep waters at the
surface of the Antarctic (François et al., 1997).
Such a change in the Antarctic would also help to
explain observations in the low-latitude ocean. For
instance, the extraction of nutrients in the Antarctic
should lower the nutrient content of the waters that
subducted into the intermediate-depth ocean and
thermocline at the equatorward margin of the
Antarctic. This would work to prevent the transfer
of nutrients from cold, deep ocean into the warmer,
upper ocean, potentially explaining an apparent
transfer of nutrients from the mid-depths to the
deep ocean during glacial times (Boyle, 1988b;
Herguera et al., 1992; Keir, 1988; Matsumoto et al.,
2002a; Sigman and Boyle, 2000).
Two possible causes for Antartic stratification
have been discussed. First, the southern hemisphere westerlies winds apparently shifted northward during glacial times (Hebbeln et al., 2002;
McCulloch et al., 2000), which should have
reduced Ekman-driven upwelling in the Antarctic
(Toggweiler et al., 1999) (“wind-shift” mechanism). Second, the vertical gradient in density, as
determined jointly by the vertical gradients in
temperature and salinity, may have been stronger
in the glacial Antarctic (“density” mechanism).
Observations
However, both of these mechanisms are currently
incomplete as explanations for the glacial
reduction in atmospheric CO2 (Keeling and
Visbeck, 2001; Sigman and Boyle, 2001). As an
example, we consider the “density” mechanism in
more detail. In the modern Antarctic, as in most
polar regions, the vertical salinity gradient tends to
stratify the upper ocean, while temperature tends
to destabilize it. Thus, either an increase in the
relative importance of salinity or a decrease in the
importance of temperature could be the fundamental trigger for the development of strong
stratification in the glacial Antarctic; once stratification begins, it will be reinforced by the
accumulation of freshwater in the upper layer.
Given the abundance of sea ice in the glacial
Antarctic (Crosta et al., 1998) and the typical
association of sea ice with freshwater release, it is
tempting to call upon sea ice as the initial driver of
stratification. However, sea ice does not represent
a source of freshwater per se, but rather a
mechanism for transporting freshwater from one
region of the Antarctic to another, with net ice
formation in the coastal Antarctic and net melting
in the open Antarctic. As a result, stratification of
the open Antarctic by the melting of sea ice might
occur at the expense of the coastal Antarctic,
where sea ice formation would make surface
waters more saline and thus more dense, leading
to overturning and the accompanying release of
CO2 to the atmosphere. Thus, we are more
encouraged by hypotheses that involve changes
in temperature (i.e., cooling of the deep ocean) as
the trigger for stratification (Gildor and
Tziperman, 2001; Gildor et al., 2002), although
these have been tested only in simple physical
models. In any case, if the evidence for stratification could be taken as overwhelming, we would
still need to determine what caused it and how
much it lowered atmospheric CO2 during glacial
times.
However, the evidence for Antarctic stratification is not overwhelming; indeed, many investigators are not convinced (e.g., Anderson et al.,
1998; Elderfield and Rickaby, 2000). The
interpretation from nitrogen isotopes of higher
nitrate utilization in the Antarctic faces a number
of apparent disagreements with other proxies of
nutrient status, in particular, the Cd/Ca (Boyle,
1988a; Keigwin and Boyle, 1989) and 13C/12C
(Charles and Fairbanks, 1988) of planktonic
foraminiferal calcite. Measurements of these
ratios in planktonic foraminifera indicate no
clear decrease in the nutrient concentration of
Antarctic surface water, while such a decrease
would have been expected on the basis of the
nitrogen isotope data. Each of the paleochemical
proxies has significant uncertainties (see Section
6.18.3.2). Nevertheless, the need to define new
interpretations of a variety of measurements to fit
519
the stratification hypothesis does not inspire
confidence.
Dissolved silicate is a major nutrient for diatom
growth because of the silica tests that these
phytoplankton precipitate. Much like nitrate and
phosphate, silicate is nearly completely depleted in
the low-latitude surface ocean but is found at
relatively high concentrations in the modern
Antarctic (Figure 5(d)). The silicon isotopic
composition of diatom microfossils implies that
there was a reduction in Antarctic silicate utilization during the last ice age (De La Rocha et al.,
1998), in contrast to the evidence for enhanced
nitrate utilization. While this may indicate a
disagreement among proxies, the alternative is
that the difference signals a real oceanographic
change. Field observations, incubations and culture studies (Franck et al., 2000; Hutchins and
Bruland, 1998; Takeda, 1998) indicate that ironreplete conditions (such as may result from the
dustiness of the ice age atmosphere (Mahowald
et al., 1999)) favor a higher nitrate/silicate ratio
in diatoms, and phytoplankton species composition changes may have reinforced this shift (Tortell
et al., 2002). Indeed, taking both the nitrogen and
silicon isotope data at face value implies that the
nitrate/silicate uptake ratio of Antarctic phytoplankton was higher during the last ice age, leading
to lower nitrate but higher silicate concentrations in
the glacial Antarctic relative to modern times
(Brzezinski et al., 2002). From the perspective of
our efforts to reconstruct the history of export
production, this possibility is problematic in that
the rain of biogenic silica out of the surface ocean
would have decreased independently from the rain
of organic carbon.
If the change in nitrate/silicate uptake ratio was
adequately great, it may have actually removed the
tendency for preferential depletion of silicate
relative to nitrate that is observed in the modern
Antarctic, possibly increasing the silicate/nitrate
ratio of Southern-source water that supplies
nutrients to the low-latitude surface (Brzezinski
et al., 2002). A greater supply of silicate to the lowlatitude ocean could have driven an increase in the
importance of silica-secreting phytoplankton (diatoms) relative to CaCO3-precipitating phytoplankton (coccolithophorids) (Matsumoto et al., 2002b);
this may explain a shift toward higher opal
accumulation rates in some equatorial regions
(Broecker et al., 2000). If increased diatom
productivity at low latitudes in the southern
hemisphere occurred at the expense of coccolithophores, the high sinking rates of diatom-derived
particulate matter may have facilitated the transport of organic matter through intermediate waters
to the deep ocean (Boyle, 1988b), again consistent
with evidence for the “nutrient deepening” during
glacial times (Boyle, 1988c, 1992; Boyle et al.,
1995; Herguera et al., 1992). In addition, a floral
520
The Biological Pump in the Past
shift away from coccolithophores to diatoms would
have also lowered the CaCO3/organic carbon rain
ratio, weakening the ocean carbonate pump. Both
the increase in the remineralization depth of
organic carbon and a decrease in the rain ratio
would have worked to lower-atmospheric CO2
(Archer and Maier-Reimer, 1994; Berger and Keir,
1984; Boyle, 1988b; Broecker and Peng, 1987;
Brzezinski et al., 2002; Dymond and Lyle,
1985; Keir and Berger, 1983; Matsumoto et al.,
2002b; Sigman et al., 1998).
In the Subantarctic, planktonic foraminiferal
Cd/Ca and the 13C/12C of diatom-bound organic
matter are consistent with an ice age state of
higher nutrient utilization (Rosenthal et al., 1997,
2000). However, the consideration of proxy
complications argue against such a change
(Elderfield and Rickaby, 2000). The significance
of the nitrogen isotope data for the glacial
Subantarctic is uncertain (François et al., 1997;
Sigman et al., 1999a). Thus, while the Subantarctic was certainly more productive during glacial
times, the history of its nutrient status is
unclear and deserves further investigation. Given
our current understanding of the role that the
Subantarctic plays in the carbon cycle, the
Subantarctic is less important in itself than in
what it indicates about the nutrient status of the
Antarctic zone at its (up-stream) southern end.
The Subantarctic also represents the gateway by
which the Antarctic affects the low-latitude ocean
and is thus central in questions regarding the
impact of Antarctic nutrient changes on lowlatitude productivity (Brzezinski et al., 2002;
Matsumoto et al., 2002b).
The Subarctic Pacific, like the Antarctic, is
characterized by year-round nonzero concentrations
of the major nutrients, nitrate and phosphate.
However, the Subarctic Pacific maintains a higher
degree of nutrient utilization (lower surface nutrient
concentrations) than the Antarctic. In the open
Subarctic Pacific, summer nitrate concentration
frequently falls below 8 mM in surface water,
despite nitrate concentrations of greater than
35 mM in the upwelling deep water below the
permanent halocline. One important difference
between the Subarctic Pacific and the Antarctic is
the much stronger halocline in the former (Reid,
1969, Talley, 1993; Warren, 1983), where the
extremely low salinity of the upper 400 m limits
the exposure of nutrient- and CO2-charged subsurface waters at the surface. It is roughly this salinitydriven stratification that has been proposed for the
glacial Antarctic. Thus, the modern Subarctic
Pacific provides something of a modern analogue
for the glacial Antarctic stratification hypothesis
(Haug et al., 1999; Morley and Hays, 1983).
The Subarctic Pacific undergoes changes in
productivity and nutrient regime over glacial/
interglacial cycles, although these changes are of
smaller amplitude than is observed for the
Southern Ocean. Opal accumulation was apparently lower during the last glacial maximum in the
western Subarctic Pacific and its marginal seas
(Gorbarenko, 1996; Haug et al., 1995; Narita et al.,
2002). This has been interpreted, in much the
same way as the Antarctic changes, as indicating
intensified stratification during glacial times
(Narita et al., 2002). In this light, the smaller
glacial/interglacial signal of this region relative to
the Antarctic is explained by the fact the modern
Subarctic Pacific is already strongly stratified.
However, the dominant feature in most sediment
records from western Subarctic Pacific and Bering
Sea is of a high-productivity event upon the
transition from the last ice age to the Holocene
(Keigwin et al., 1992; Nakatsuka et al., 1995), for
which there is as yet no well-supported explanation. Moreover, the eastern Subarctic Pacific
seems to lack a glacial/interglacial signal, instead
being characterized by abrupt events of opal
deposition that are not clearly linked to global
climate change (McDonald et al., 1999).
6.18.5 SUMMARY AND CURRENT OPINION
While the early focus of hypotheses for glacial/interglacial CO2 change was on the biological
pump (Broecker, 1982a,b), a number of alternative
mechanisms, mostly involving the calcium carbonate budget, attracted increasing attention
during the late 1980s and the 1990s (Archer and
Maier-Reimer, 1994; Boyle, 1988b; Broecker and
Peng, 1987; Opdyke and Walker, 1992). As
problems have been recognized with these mechanisms as the sole driver of glacial/interglacial CO2
change (Archer and Maier-Raimer, 1994; Sigman
et al., 1998), the biological pump is receiving
attention once again. Hypotheses exist for both
low- and high-latitude changes in the biological
pump. Their strengths, weaknesses, and central
questions are now fairly clear.
Currently, the most popular low-latitude
hypotheses depend largely on whether the nitrogen cycle alone can drive a large-scale change in
low-latitude export production. As of the early
2000s, while this question has not yet been directly
addressed, no new observations have arisen that
would overturn the traditional view that the nitrogen cycle could have had only a limited effect
without cooperation from the phosphorus cycle
and that the phosphorus cycle could not have been
adequately dynamic over glacial/interglacial
cycles (e.g., Froelich et al., 1982; Haug et al.,
1998; Redfield, 1942, 1958; Redfield et al., 1963;
Ruttenberg, 1993; Sanudo-Wilhelmy et al., 2001).
Thus, while the low-latitude biological pump has
by no means been eliminated as the driver of
glacial/interglacial cycles, our attention is focused
on the polar ocean, the Antarctic in particular.
Summary and Current Opinion
521
Figure 11 A 6 Myr biogenic opal record from Ocean Drilling Program Site 882 in the western Subarctic Pacific
(508210 N, 1678350 E; water depth 3,244 m; figure from Haug et al. (1999)), showing an approximately fourfold
decrease in opal accumulation rate at 2.73 Myr ago (top panel). This abrupt drop in opal occurred at isotope stage G6
(the former isotope stage 110) synchronously with the massive onset of ice rafted debris, as indicated by the increase
in magnetic susceptibility (bottom panel). Maxima in opal accumulation during the last 2.73 Myr are generally linked
to interglacial times or deglaciations, as reported by Keigwin et al. (1992) for the last deglaciation. Since silicate is
supplied by the exposure of nutrient-rich deep water, the high opal flux rates of the mid-Pliocene Subarctic Pacific
require a rate of deep-water exposure of a magnitude similar to that observed in the modern Antarctic. Because
silicate consumption is nearly complete in the modern Subarctic Pacific (Figure 5(d)), the sharp drop in opal flux at
2.73 Myr ago cannot be attributed simply to a decrease in the completeness of silicate consumption. Rather, it must
record a decrease in the rate of exposure of silicate-rich deep water at the Subarctic Pacific surface. Sediment d 15N
(15N/14Nsample/15N/14Nair 2 1) £ 1,000) increases concurrently with the sharp drop in opal accumulation at the
event 2.73 Myr ago (from ,3‰ to 5‰, upper panel). This shift toward higher d 15N suggests an increase in nitrate
utilization (Altabet and Francois, 1994a), providing additional evidence that the decrease in opal flux resulted from a
decrease in the exposure of nutrient-rich deep water, which lowered the nitrate supply and thus forced more complete
nitrate consumption while decreasing the silicate available for opal production (reproduced by permission of Nature
Publishing Group from Nature 401, 779– 782).
We have discussed three rough categories of
Antarctic change that might drive a decline in CO2
during ice ages: (i) increased biological productivity, as in response to increased iron supply,
(ii) vertical stratification of the upper ocean, and
(iii) gas exchange limitation by ice cover. The
productivity hypothesis is not supported by data,
which suggest a smaller biogenic rain in the
glacial Antarctic, although it remains possible that
our proxies are telling us only about a reduction in
the rain of bulk biogenic material (i.e., diatomderived opal) and are leading us to overlook an
increase in “export production,” the rain of
organic carbon out of the surface ocean. The gas
exchange hypothesis remains possible but is
difficult to test and would seem to require extreme
conditions to be the sole cause for the observed
amplitude of CO2 change. The stratification
hypothesis is supported by some data but conflicts
with other data, or at least with the traditional
interpretations of those data. While there are
major concerns about each of the polar hypotheses, these hypotheses have recently been
strengthened by the realization that they would
have impacted the lower-latitude ocean and the
CaCO3 budget in ways that should have strengthened their overall effect on atmospheric CO2 (e.g.,
Brzezinski et al., 2002; Matsumoto et al., 2002b;
Toggweiler, 1999).
As is clear from the treatment above, the
stratification hypothesis has particular resonance
with the authors of this chapter. This is driven
522
The Biological Pump in the Past
largely by the conceptual and observational links
we observe between two nutrient-bearing polar
ocean regions, the Antarctic and the Subarctic
Pacific. As described above, there is a strong
analogy between the Subarctic Pacific that we
observe today and the glacial-age Antarctic that is
posed by the stratification hypothesis. Moreover,
at a major cooling event 2.7 Myr ago, it appears
that the Subarctic Pacific underwent the transition
to its current salinity-stratified condition
(Figure 11) (Haug et al., 1999), just as stratification has been hypothesized for the Antarctic on the
transitions to ice age conditions. Ongoing work on
other intervals of the climate record seems to
support the existence of a strong generalized link
between climate cooling and polar ocean stratification. The physical mechanism for this link,
however, is still unresolved.
To these authors, there is no infallible recipe for
the investigation of the biological pump in the past.
Nevertheless, there are glaring uncertainties that
require our attention, both in our concepts and our
tools. The quantitative effects of the low latitude
and polar ocean on atmospheric CO2 are still a
matter of debate (Broecker et al., 1999; Toggweiler
et al., 2003a,b). Until our understanding of the
modern ocean and carbon cycle has progressed to
the degree that such major conceptual questions
can be resolved, it will be extremely difficult to
generate any consensus on the role of the biological
pump in driving carbon dioxide change. After
optimistic discussion of “proxy calibration” in the
1990s, we now recognize that the sedimentary and
geochemical parameters we use to study the
history of the biological pump are the product of
multiple environmental variables, requiring that
we think deeply about their significance in each
case that they are applied. The quantity of work on
the history of the biological pump has increased
markedly over the last decade; a concerted effort to
review, evaluate, and synthesize this information
would greatly improve its usefulness. At the same
time, we must continue to support a vanguard in
search of untapped constraints. We have emphasized above that the biological pump is not strictly
a biological phenomenon but rather results from
the interaction of ocean biology, chemistry, and
physics; as such, progress in its study is tied to our
understanding of the history of the ocean in
general, including its physical circulation and
fundamental conditions (Adkins et al., 2002).
ACKNOWLEDGMENTS
DMS is supported by the US NSF through
grants OCE-9981479, OCE-0081686, DEB0083566 (to Simon Levin), and OCE-0136449,
and by British Petroleum and Ford Motor
Company through the Princeton Carbon Mitigation Initiative. This chapter is dedicated to the
memory of David S. Sigman, the only biochemist
who would have tried to read it.
REFERENCES
Adkins J. F., McIntyre K., and Schrag D. P. (2002) The salinity,
temperature, and delta O-18 of the glacial deep ocean.
Science 298(5599), 1769–1773.
Altabet M. A. (2001) Nitrogen isotopic evidence for micronutrient control of fractional NO2
3 utilization in the
equatorial Pacific. Limnol. Oceanogr. 46(2), 368–380.
Altabet M. A. and Francois R. (1994a) Sedimentary nitrogen
isotopic ratio as a recorder for surface ocean nitrate
utilization. Global Biogeochem. Cycles 8(1), 103– 116.
Altabet M. A. and Francois R. (1994b) The use of nitrogen
isotopic ratio for reconstruction of past changes in surface
ocean nutrient utilization. In Carbon Cycling in the Glacial
Ocean: Constraints on the Ocean’s Role in Global Change
(eds. R. Zhan, M. Kaminski, L. Labeyrie, and T. F.
Pederson). Springer, Berlin, Heidelberg, New York vol.
17, pp. 281–306.
Altabet M. A., Francois R., Murray D. W., and Prell W. L.
(1995) Climate-related variations in denitrification in the
Arabian Sea from sediment 15N/14N ratios. Nature 373,
506–509.
Altabet M. A., Higginson M. J., and Murray D. W. (2002) The
effect of millennial-scale changes in Arabian Sea denitrification on atmospheric CO2. Nature 415(6868), 159–162.
Anderson D. M. and Prell W. L. (1993) A 300 kyr record of
upwelling off Oman during the Late Quaternary: evidence of
the Asian southwest monsoon. Paleoceanography 8,
193–208.
Anderson R., Kumar N., Mortlock R., Froelich P., Kubik P.,
Dittrich-Hannen B., and Suter M. (1998) Late-Quaternary
changes in productivity of the Southern Ocean. J. Mar. Sys.
17, 497 –514.
Anderson R. F., Lehuray A. P., Fleisher M. Q., and Murray
J. W. (1989) Uranium deposition in Saanich Inlet sediments,
Vancouver Island. Geochim. Cosmochim. Acta 53(9),
2205–2213.
Anderson R. F., Lao Y., Broecker W. S., Trumbore S. E.,
Hofmann H. J., and Wolfi W. (1990) Boundary scavenging
in the Pacific Ocean: a comparison of 10Be and 231Pa. Earth
Planet. Sci. Lett. 96, 287 –304.
Archer D. and Maier-Reimer E. (1994) Effect of deep-sea
sedimentary calcite preservation on atmospheric CO2
concentration. Nature 367, 260 –263.
Archer D., Winguth A., Lea D., and Mahowald N. (2000a)
What caused the glacial/interglacial atmospheric pCO2
cycles? Rev. Geophys. 38(2), 159–189.
Archer D. E., Eshel G., Winguth A., Broecker W. S.,
Pierrehumbert R., Tobis M., and Jacob R. (2000b)
Atmospheric pCO2 sensitivity to the biological pump in
the ocean. Global Biogeochem. Cycles 14(4), 1219–1230.
Arrigo K. R., Robinson D. H., Worthen D. L., Dunbar R. B.,
DiTullio G. R., VanWoert M., and Lizotte M. P. (1999)
Phytoplankton community structure and the drawdown of
nutrients and CO2 in the Southern Ocean. Science
283(5400), 365 –367.
Bacon M. P. (1984) Glacial to interglacial changes in carbonate
and clay sedimentation in the Atlantic Ocean estimated from
Th-230 measurements. Isotope Geosci. 2, 97–111.
Bareille G., Labracherie M., Mortlock R. A., Maier-Reimer E.,
and Froelich P. N. (1998) A test of (Ge/Si)opal
as a paleorecorder of (Ge/Si)seawater. Geology 26(2),
179–182.
Barnola J. M., Raynaud D., Korotkevich Y. S., and Lorius C.
(1987) Vostok ice core provides 160,000-year record of
atmospheric CO2. Nature 329, 408414.
Bassinot F. C., Labeyrie L. D., Vincent E., Quidellerur X.,
Shackleton N. J., and Lancelot Y. (1994) The astronomical
References
theory of climate and the age of the Brunhes-Matuyama
magnetic reversal. Earth Planet. Sci. Lett. 126, 91–108.
Bauch D., Carstens J., and Wefer G. (1997) Oxygen isotope
composition of living Neogloboquadrina pachyderma (sin.)
in the Arctic Ocean. Earth Planet. Sci. Lett. 146(1–2),
47–58.
Beaufort L. (2000) Dynamics of the monsoon in the equatorial
Indian Ocean over the last 260,000 years. Quat Int. 31,
13–18.
Berger A., Imbrie J., Hays J., Kukla G., and Saltzman B.
(1984). Milankovitch and Climate, Reidel, Boston vol. 1,
493pp.
Berger W. H. and Keir R. S. (1984) Glacial-Holocene changes
in atmospheric CO2 and the deep-sea record. In Climate
Processes and Climate Sensitivity. Geophyisical Monograph
29 (eds. J. E. Hansen and T. Takahashi). American
Geophysical Union, Washington DC, pp. 337 –351.
Berger W. H. and Lange C. B. (1998) Silica depletion in the
thermocline of the glacial North Pacific: corollaries and
implications. Deep-Sea Res. II 45(8–9), 1885– 1904.
Boyle E. A. (1988a) Cadmium: chemical tracer of deepwater
paleoceanography. Paleoceanography 3(4), 471 –489.
Boyle E. A. (1988b) The role of vertical chemical fractionation
in controlling late quaternary atmospheric carbon dioxide.
J. Geophys. Res. 93(C12), 15701–15714.
Boyle E. A. (1988c) Vertical oceanic nutrient fractionation and
glacial/interglacial CO2 cycles. Nature 331, 55– 56.
Boyle E. A. (1992) Cadmium and d 13C paleochemical ocean
distributions during the stage 2 glacial maximum. Ann. Rev.
Earth Planet. Sci. 20, 245– 287.
Boyle E. A. (1998) Pumping iron makes thinner diatoms.
Nature 393, 733–734.
Boyle E. A. and Keigwin L. D. (1982) Deep circulation of the
North Atlantic for the last 200,000 years: geochemical
evidence. Science 218, 784 –787.
Boyle E. A., Labeyrie L., and Duplessy J.-C. (1995) Calcitic
foraminiferal data confirmed by cadmium in aragonitic
Hoeglundina: application to the last glacial maximum in the
northern Indian Ocean. Paleoceanography 10(5), 881–900.
Broecker W. S. (1981) Glacial to interglacial changes in ocean
and atmospheric chemistry. In Climatic Variations and
Variability: Facts and Theory (ed. A. Berger). Kluwer,
Boston, pp. 111–121.
Broecker W. S. (1982a) Glacial to interglacial changes in ocean
chemistry. Progr. Oceanogr. 2, 151–197.
Broecker W. S. (1982b) Ocean chemistry during glacial time.
Geochim. Cosmochim. Acta 46, 1689–1706.
Broecker W. S. and Henderson G. M. (1998) The sequence of
events surrounding Termination II and their implications for
the cause of glacial– interglacial CO2 changes. Paleoceanography 13(4), 352 –364.
Broecker W. S. and Maier-Reimer E. (1992) The influence of
air and sea exchange on the carbon isotope distribution in the
sea. Global Biogeochem. Cycles 6(3), 315 –320.
Broecker W. S. and Peng T.-H. (1982) Tracers in the Sea.
Eldigio Press, Palisades, New York.
Broecker W. S. and Peng T.-H. (1987) The role of CaCO3
compensation in the glacial to interglacial atmospheric CO2
change. Global Biogeochem. Cycles 1(1), 15 –29.
Broecker W. S., Peacock S. L., Walker S., Weiss R., Fahrbach
E., Schroeder M., Mikolajewicz U., Heinze C., Key R., Peng
T. H., and Rubin S. (1998) How much deep water is formed
in the Southern Ocean? J. Geophys. Res.: Oceans 103(C8),
15833– 15843.
Broecker W. S., Lynch-Stieglitz J., Archer D., Hofmann M.,
Maier-Reimer E., Marchal O., Stocker T., and Gruber N.
(1999) How strong is the Harvardton-Bear constraint?
Global Biogeochem. Cycles 13(4), 817–820.
Broecker W. S., Clark E., Stiegllitz J. L., Beck W., Stott L. D.,
Hajdas I., and Bonani G. (2000) Late glacial diatom
accumulation at 98 S in the Indian Ocean. Paleoceanography
15, 348 –352.
523
Brzezinski M. A., Pride C. J., Franck V. M., Sigman D. M.,
Sarmiento J. L., Matsumoto K., Gruber N., Rau G. H., and
Coale K. H. (2002) A switch from Si(OH)4 to NO2
3 depletion
in the glacial Southern Ocean. Geophys. Res. Lett. 29(12),
10.1029/12001GL014349.
Canfield D. E. (1994) Factors influencing organic carbon
preservation in marine sediments. Deep-Sea Res. 114,
315 –329.
Charles C. D. and Fairbanks R. G. (1988) Glacial to interglacial
changes in the isotopic gradients of southern Ocean surface
water. In Geological History of the Polar Oceans: Arctic
versus Antarctic (eds. U. Bleil and J. Thiede). Kluwer
Academic, pp. 519– 538.
Chase Z., Anderson R. F., and Fleisher M. Q. (2001) Evidence
from authigenic uranium for increased productivity of the
glacial Subantarctic Ocean. Paleoceanography 16(5),
468 –478.
Chase Z., Anderson R. F., Fleisher M. Q., and Kubik P. W.
(2003a) Accumulation of biogenic and lithogenic material in
the Pacific sector of the Southern Ocean during the past
40,000 years. Deep-Sea Res. II 50(3–4), 739–768.
Chase Z., Anderson R. F., Fleisher M. Q., and Kubik P. W.
(2003b) Scavenging of 230Th, 231Pa and 10Be in the Southern
Ocean (SW Pacific sector): the importance of particle flux,
particle composition and advection. Deep-Sea Res. II 50
(3–4), 739–768.
Clemens S. C., Prell W. L., Murray D. W., Shimmield G., and
Weedon G. (1991) Forcing mechanisms of the Indian Ocean
monsoon. Nature 353, 720 –725.
Codispoti L. A. (1995) Biogeochemical cycles: Is the ocean
losing nitrate? Nature 376(6543), 724.
Conkright M., Levitus S., and Boyer T. (1994) World Ocean
Atlas 1994, vol. 1: Nutrients. US Department of Commerce,
Washington DC, pp. 16.
Crosta X., Pichon J.-J., and Burckle L. H. (1998) Application of
modern analogue technique to marine Antarctic diatoms:
reconstruction of maximum sea-ice extent at the Last Glacial
Maximum. Paleoceanography 13(3), 284–297.
Crusius J. and Thomson J. (2000) Comparative behavior of
authigenic Re, U, and Mo during reoxidation and subsequent
long-term burial in marine sediments. Geochim. Cosmochim.
Acta 64(13), 2233–2242.
Crusius J., Calvert S., Pedersen T., and Sage D. (1996)
Rhenium and molybdenum enrichnments in sediments as
indicators of oxic, suboxic and sulfidic conditions of
deposition. Earth Planet. Sci. Lett. 145(1–4), 65 –78.
Dean W. E., Gardner J. V., and Piper D. Z. (1997) Inorganic
geochemical indicators of glacial–interglacial changes in
productivity and anoxia on the California continental
margin. Geochim. Cosmochim. Acta 61(21), 4507–4518.
Dehairs F., Stroobants N., and Goeyens L. (1991) Suspended
barite as a tracer of biological activity in the Southern Ocean.
Mar. Chem. 35, 399 –410.
De La Rocha C. L., Brzezinski M. A., and DeNiro M. J. (1997)
Fractionation of silicon isotopes by marine diatoms during
biogenic silica formation. Geochim. Cosmochim. Acta 61,
5051–5056.
De La Rocha C. L., Brzezinski M. A., DeNiro M. J., and
Shemesh A. (1998) Silicon-isotope composition of diatoms
as an indicator of past oceanic change. Nature 395(6703),
680 –683.
Deutsch C., Gruber N., Key R. M., Sarmiento J. L., and
Ganaschaud A. (2001) Denitrification and N2 fixation in the
Pacific Ocean. Global Biogeochem. Cycles 15(2), 483 –506.
DiTullio G. R., Grebmeier J. M., Arrigo K. R., Lizotte M. P.,
Robinson D. H., Leventer A., Barry J. P., VanWoert M. L.,
and Dunbar R. B. (2000) Rapid and early export of
Phaeocystis antarctica blooms in the Ross Sea, Antarctica.
Nature 404, 595–598.
Duplessy J. C. (1982) Glacial to interglacial contrasts in the
northern Indian Ocean. Nature 295, 494– 498.
Dymond J. and Lyle M. (1985) Flux comparisons between
sediments and sediment traps in the eastern tropical Pacific:
524
The Biological Pump in the Past
implications for atmospheric CO2 variations during the
Pleistocene. Limnol. Oceanogr. 30(4), 699–712.
Dymond J., Suess E., and Lyle M. (1992) Barium in deep-sea
sediment: a geochemical proxy for paleoproductivity.
Paleoceanography 7(2), 163–181.
Elderfield H. and Rickaby R. E. M. (2000) Oceanic Cd/P ratio
and nutrient utilization in the glacial Southern Ocean. Nature
405, 305–310.
Elser J. J., Sterner R. W., Gorokhova E., Fagan W. F., Markow
T. A., Cotner J. B., Harrison J. F., Hobbie S. E., Odell G. M.,
and Weider L. J. (2000) Biological stoichiometry from genes
to ecosystems. Ecol. Lett. 3, 540– 550.
Emeis K.-C., Anderson D. M., Doose H., Kroon D., and
Schulz-Bull D. E. (1995) Sea-surface temperatures and the
history of monsoon upwelling in the NW Arabian Sea during
the last 500,000 yr. Quat. Res. 43, 355–361.
Emmer E. and Thunell R. C. (2000) Nitrogen isotope variations
in Santa Barbara Basin sediments: implications for denitrification in the eastern tropical North Pacific during the last
50,000 years. Paleoceanography 15(4), 377 –387.
Fairbanks R. G. (1989) A 17,000-year glacio-eustatic sea level
record: influence of glacial melting rates on the Younger
Dryas event and deep-ocean circulation. Nature 342,
637 –642.
Falkowski P. G. (1997) Evolution of the nitrogen cycle and its
influence on the biological sequestration of CO2 in the
ocean. Nature 387(6630), 272–275.
Falkowski P. G. (2000) Rationalizing elemental ratios in
unicellular algae. J. Phycol. 36, 3–6.
Farrell J. W., Pedersen T. F., Calvert S. E., and Nielsen B.
(1995) Glacial–interglacial changes in nutrient utilization in
the equatorial Pacific Ocean. Nature 377(6549), 514 –517.
Fischer G. and Wefer G. (1999) Use of Proxies in
Paleoceanography: Examples from the South Atlantic (eds.
G. Fischer and G. Wefer). Springer, Berlin, 727pp.
Fontugne M. R. and Duplessy J.-C. (1986) Variations of the
monsoon regime during the upper Quaternary: evidence
from carbon isotopic record of organic matter in North
Indian Ocean sediment cores. Palaeogeogr. Paleoclimatol.
Palaeoecol. 56, 69–88.
FranceLanord C. and Derry L. A. (1997) Organic carbon burial
forcing of the carbon cycle from Himalayan erosion. Nature
390(6655), 65–67.
Franck V. M., Brzezinski M. A., Coale K. H., and Nelson D. M.
(2000) Iron and silicic acid availability regulate Si uptake in
the Pacific Sector of the Southern Ocean. Deep-Sea Res. II
47, 3315–3338.
François R., Bacon M. P., and Suman D. O. (1990) Thorium230 profiling in deep-sea sediments: high-resolution records
of flux and dissolution of carbonate in the equatorial Atlantic
during the last 24,000 years. Paleoceanography 5, 761 –787.
François R. F., Altabet M. A., Yu E.-F., Sigman D. M., Bacon
M. P., Frank M., Bohrmann G., Bareille G., and Labeyrie
L. D. (1997) Water column stratification in the Southern
Ocean contributed to the lowering of glacial atmospheric
CO2. Nature 389, 929–935.
Frank M., Gersonde R., van der Loeff M.R., Bohrmann G.,
Nurnberg C. C., Kubik, P. W., Suter M., and Mangini A.
(2000) Similar glacial and interglacial export bioproductivity in the Atlantic sector of the Southern Ocean: multiproxy
evidence and implications for glacial atmospheric CO2.
Paleoceanography 15(6), 642– 658, 2000PA000497.
Frew R. D. and Hunter K. A. (1992) Influence of Southern
Ocean waters on the cadmium-phosphate properties of the
global ocean. Nature 360(6400), 144–146.
Froelich P. N. and Andreae M. O. (1981) The marine
geochemistry of germanium: ekasilicon. Science 213,
205 –207.
Froelich P. N., Bender M. L., Luedtke N. A., Heath G. R., and
DeVries T. (1982) The marine phosphorus cycle. Am. J. Sci.
282, 474–511.
Ganeshram R. S. and Pedersen T. F. (1998) Glacial –
interglacial variability in upwelling and bioproductivity off
NW Mexico: implications for quaternary paleoclimate.
Paleoceanography 13(6), 634–645.
Ganeshram R. S., Pedersen T. F., Calvert S. E., and Murray
J. W. (1995) Large changes in oceanic nutrient inventories
from glacial to interglacial periods. Nature 376(6543),
755–758.
Ganeshram R. S., Pedersen T. F., Calvert S. E., McNeill G. W.,
and Fontugne M. R. (2000) Glacial–interglacial variability in
denitrification in the world’s oceans: causes and consequences. Paleoceanography 15(4), 361 –376.
Ganeshram R. S., Pedersen T. F., Calvert S. E., and Francois R.
(2002) Reduced nitrogen fixation in the glacial ocean
inferred from changes in marine nitrogen and phosphorus
inventories. Nature 415, 156–159.
Gildor H. and Tziperman E. (2001) Physical mechanisms
behind biogeochemical glacial– interglacial CO2 variations.
Geophys. Res. Lett. 28, 2421–2424.
Gildor H., Tziperman E., and Toggweiler J. R. (2002) Sea ice
switch mechanism and glacial–interglacial CO2 variations.
Global Biogeochem. Cycles 16(3) 10.1029/2001GB001446.
Gingele F. X., Zabel M., Kasten S., Bonn W. J., and Nurnberg
C. C. (1999) Biogenic barium as a proxy for paleoproductivity: methods and limitations of application. In Use of
Proxies in Paleoceanography: Examples from the South
Atlantic (eds. G. Fischer and G. Wefer). Springer, Berlin,
pp. 345 –364.
Gorbarenko S. A. (1996) Stable isotope and lithologic evidence
of late-Glacial and Holocene oceanography of the northwestern Pacific and its marginal seas. Quat. Res. 46,
230–250.
Gordon A. L., Taylor H. W., and Georgi D. T. (1977) Antarctic
oceanographic zonation. In Polar Oceans (ed. M. J. Dunbar).
Arctic Institute of North America, pp. 219 –225.
Gruber N. and Sarmiento J. L. (1997) Global patterns of marine
nitrogen fixation and denitrification. Global Biogeochem.
Cycles 11, 235–266.
Hastings D. W., Emerson S. R., and Mix A. C. (1996)
Vanadium in foraminiferal calcite as a tracer for changes in
the areal extent of reducing conditions. Paleoceanography
11(6), 666–678.
Haug G. H., Maslin M. A., Sarnthein M., Stax R., and
Tiedemann R. (1995) Evolution of northwest Pacific
sedimentation patterns since 6 Ma: site 882. Proc. ODP.
Sci. Results 145, 293–315.
Haug G. H., Pedersen T. F., Sigman D. M., Calvert S. E.,
Nielsen B., and Peterson L. C. (1998) Glacial/interglacial
variations in productivity and nitrogen fixation in the
Cariaco Basin during the last 550 ka. Paleoceanography
13(5), 427–432.
Haug G. H., Sigman D. M., Tiedemann R., Pedersen T. F., and
Sarnthein M. (1999) Onset of permanent stratification in the
Subarctic Pacific Ocean. Nature 40, 779 –782.
Hays J. D., Imbrie J., and Shackleton N. J. (1976) Variations in
the Earth’s orbit: pacemaker of the ice ages. Science 194,
1121–1132.
Hebbeln D., Marchant M., and Wefer G. (2002) Paleoproductivity in the southern Peru-Chile Current through the last 33
000 yr. Mar. Geol. 186(3–4), 487– 504.
Heinze H.-M. and Wefer G. (1992) The history of coastal
upwelling off Peru over the past 650,000 years. In Upwelling
Systems: Evolution Since the Early Miocene (eds. C. P.
Summerhayes, W. L. Prell, and K.-C. Emeis). Geological
Society of London, vol. 64, pp. 451 –462.
Herbert T. D., Schuffert J. D., Andreasen D., Heusser L., Lyle
M., Mix A., Ravelo A. C., Stott L. D., and Herguera J. C.
(2001) Collapse of the California current during glacial
maxima linked to climate change on land. Science 293,
71–76.
Herguera J. C. (1992) Deep-sea benthic foraminifera and
biogenic opal—glacial to postglacial productivity changes in
the western equatorial Pacific. Mar. Micropaleonthol. 19,
79–98.
References
Herguera J. C. and Berger W. H. (1991) Paleoproductivity from
benthic foraminifera abundance—glacial to postglacial
change in the west-equatorial Pacific. Geology 19(12),
1173– 1176.
Herguera J. C. and Berger W. H. (1994) Glacial to postglacial
drop in productivity in the western equatorial Pacific—
mixing rate versus nutrient concentrations. Geology 22(7),
629–632.
Herguera J. C., Jansen E., and Berger W. H. (1992) Evidence
for a bathyal front at 2,000 m depth in the glacial Pacific,
based on a depth transect on Ontong Java Plateau.
Paleoceanography 7(3), 273–288.
Hinrichs K. U., Schneider R. R., Muller P. J., and Rullkotter J.
(1999) A biomarker perspective on paleoproductivity
variations in two Late Quaternary sediment sections from
the Southeast Atlantic Ocean. Org. Geochem. 30(5),
341–366.
Hofmann M., Broecker W. S., and Lynch-Stieglitz J. (1999)
Influence of a [CO2(aq)] dependent biological C-isotope
fractionation on glacial C-13/C-12 ratios in the ocean.
Global Biogeochem. Cycles 13(4), 873–883.
Holmen K. (1992) The global carbon cycle. In Global
Biogeochemical Cycles (eds. S. S. Butcher, R. J. Charlson,
G. H. Orians, and G. V. Wolfe). Academic Press, New York,
pp. 239 –262.
Holmes M. E., Schneider R. R., Muller P. J., Segl M., and
Wefer G. (1997) Reconstruction of past nutrient utilization
in the eastern Angola Basin based on sedimentary 15N/14N
ratios. Paleoceanography 12(4), 604 –614.
Huang C.-Y., Liew P.-M., Zhao M., Chang T.-C., Kuo C.-M.,
Chen M.-T., Wang C.-H., and Zheng L.-F. (1997) Deep sea
and lake records of the Southeast Asian paleomonsoons for
the last 25 thousand years. Earth Planet. Sci. Lett. 146,
59–72.
Hutchins D. A. and Bruland K. W. (1998) Iron-limited diatom
growth and Si : N uptake ratios in a coastal upwelling
regime. Nature 393, 561– 564.
Janecek T. R. (2001) Data report: Late Pleistocene biogenic
opal data for Leg 177 Sites 1093 and 1094. Proc. Ocean
Drilling Program, Sci. Results 177, 1 –5.
Karl D., Letelier R., Tupas L., Dore J., Christian J., and Hebel
D. (1997) The role of nitrogen fixation in biogeochemical
cycling in the subtropical North Pacific Ocean. Nature
388(6642), 533 –538.
Kawahata H., Suzuki A., and Ahagon N. (1998) Biogenic
sediments in the West Caroline Basin and the western
equatorial Pacific during the last 330,000 years. Mar. Geol.
149, 155 –176.
Kawahata H., Ohkushi K. I., and Hatakeyama Y. (1999)
Comparative Late Pleistocene paleoceanographic changes in
the mid latitude Boreal and Austral Western Pacific.
J. Oceanogr. 55, 747 –761.
Keeling R. F. and Stephens B. B. (2001) Antarctic sea ice and
the control of Pleistocene climate instability. Paleoceanography 16(1), 112 –131.
Keeling R. F. and Visbeck M. (2001) Palaeoceanography:
Antarctic stratification and glacial CO2. Nature 412,
605–606.
Keigwin L. D. and Boyle E. A. (1989) Late quaternary
paleochemistry of high-latitude surface waters. Paleogeogr.
Palaeoclimatol. Paleoecol. 73, 85–106.
Keigwin L. D., Jones G. A., and Froelich P. N. (1992) A 15,000
year paleoenvironmental record from Meiji Seamount, far
northwestern Pacific. Earth Planet. Sci. Lett. 111, 425–440.
Keil R. G. and Cowie G. L. (1999) Organic matter preservation
through the oxygen-deficient zone of the NE Arabian Sea as
discerned by organic carbon: mineral surface area ratios.
Mar. Geol. 161(1), 13 –22.
Keir R. and Berger W. (1983) Atmospheric CO2 content in the
last 120,000 years: the phosphate extraction model.
J. Geophys. Res. 88, 6027–6038.
Keir R. S. (1988) On the late Pleistocene ocean geochemistry
and circulation. Paleoceanography 3, 413 –445.
525
Keller K. and Morel F. M. M. (1999) A model of carbon
isotopic fractionation and active carbon uptake in phytoplankton. Mar. Ecol. Prog. Ser. 182, 295–298.
Kohfeld K. E., Fairbanks R. G., Smith S. L., and Walsh I. D.
(1996) Neogloboquadrina pachyderma (sinistral coiling) as
paleoceanographic tracers in polar oceans: evidence from
northeast water polynya plankton tows, sediment traps, and
surface sediments. Paleoceanography 11(6), 679–699.
Kumar N., Anderson R. F., Mortlock R. A., Froelich P. N.,
Kubik P., Dittrich-Hannen B., and Suter M. (1995) Increased
biological productivity and export production in the glacial
Southern Ocean. Nature 378, 675 –680.
Leuenberger M., Siegenthaler U., and Langway C. C. (1992)
Carbon isotope composition of atmospheric CO2 during the
last Ice Age from an Antarctic ice core. Nature 357(6378),
488 –490.
Loubere P. (1999) A multiproxy reconstruction of biological
productivity and oceanography in the eastern equatorial
Pacific for the past 30,000 years. Mar. Micropaleontol. 37,
173 –198.
Lourey M. J., Trull T. W., and Sigman D. M. (2003) An
unexpected decrease of d 15 N of surface and deep organic
nitrogen accompanying Southern Ocean seasonal nitrate
depletion. Global Biogeochem. Cycles (in press).
Lyle M. (1988) Climatically forced organic carbon burial in the
equatorial Atlantic and Pacific Oceans. Nature 335,
529 –532.
Lyle M., Murray D., Finney B., Dymond J., Robbins J., and
Brookforce K. (1988) The record of Last Pleistocene
biogenic sedimentation in the eastern tropical Pacific
Ocean. Paleoceanography 3, 39–59.
Mahowald N., Kohfeld K., Hansson M., Balkanski Y., Harrison
S. P., Prentice I. C., Schulz M., and Rodhe H. (1999) Dust
sources and deposition during the last glacial maximum and
current climate: a comparison of model results with
paleodata from ice cores and marine sediments. J. Geophys.
Res.: Atmos. 104, 15895–15916.
Manighetti B. and McCave I. (1995) Depositional fluxes and
paleoproductivity and ice rafting in the NE Atlantic over the
past 30 ka. Paleoceanography 10, 579–592.
Marcantonio F., Anderson R. F., Stute M., Kumar N., Schlosser
P., and A M. (1996) Extraterrestrial He-3 as a tracer of
marine sediment transport and accumulation. Nature
383(6602), 705–707.
Marcantonio F., Anderson R., Higgins S., Fleisher M., Stute
M., and Schlosser P. (2001a) Abrupt intensification of the
SW Indian Ocean monsoon during the last delaciation:
constraints from Th, Pa and He isotopes. Earth Planet. Sci.
Lett. 184, 505 –514.
Marcantonio F., Anderson R., Higgins S., Stute M., and
Schlosser P. (2001b) Sediment focussing in the central
equatorial Pacific Ocean. Paleoceanography 16, 260–267.
Marchitto T. M., Curry W. B., and Oppo D. W. (1998)
Millennial-scale changes in North Atlantic circulation since
the last glaciation. Nature 393(6685), 557–561.
Marino B. D. and McElroy M. B. (1991) Isotopic composition
fo atmospheric CO2 inferred from carbon in C4 plant
cellulose. Nature 349, 127– 131.
Martin J. H. (1990) Glacial–interglacial CO2 change: the iron
hypothesis. Paleoceanography 5, 1–13.
Martin J. H., Fitzwater S. E., and Gordon R. M. (1990) Iron
deficiency limits growth in Antarctic waters. Global
Biogeochem. Cycles 4, 5–12.
Martinez P. H., Bertrand P. H., Bouloubassi I., Bareille G.,
Shimmield G., Vautravers B., Grousset F., Guichard S.,
Ternois Y., and Sicre M.-A. (1996) An integrated view
of inorganic and organic biogeochemical indicators of
paleoproductivity changes in a coastal upwelling area.
Org. Geochem. 24, 411 –420.
Matsumoto K., Oba T., Lynch-Stieglitz J., and Yamamoto H.
(2002a) Interior hydrography and circulation of the glacial
Pacific Ocean. Quat. Sci. Rev. 21, 1693–1704.
526
The Biological Pump in the Past
Matsumoto K., Sarmiento J. L., and Brzezinski M. A. (2002b)
Silicic acid “leakage” from the Southern Ocean as a possible
mechanism for explaining glacial atmospheric pCO2. Global
Biogeochem. Cycles 16 10.1029/2001GB001442.
McCorkle D. C., Martin P. A., Lea D. W., and Klinkhammer
G. P. (1995) Evidence of a dissolution effect on benthic
foraminiferal shell chemistry: delta C-13, Cd/Ca, Ba/Ca, and
Sr/Ca results from the Ontong Java Plateau. Paleoceanography 10(4), 699–714.
McCulloch R. D., Bentley M. J., Purves R. S., Hulton N. R. J.,
Sugden D. E., and Clapperton C. M. (2000) Climate
inferences from glacial and palaeoecological evidence at
the last glacial termination, southern South America. J. Quat.
Sci. 15, 409 –417.
McDonald D., Pederson T., and Crusius J. (1999) Multiple Late
Quaternary episodes of exceptional diatom production in the
Gulf of Alaska. Deep-Sea Res. II 46, 2993–3017.
McElroy M. B. (1983) Marine biological controls on
atmospheric CO2 and climate. Nature 302, 328 –329.
McIntyre A. and Molfino B. (1996) Forcing of Atlantic
equatorial and subpolar millennial cycles by precession.
Science 274(5294), 1867–1870.
Mitchell B. G., Brody E. A., Holm-Hansen O., McClain C., and
Bishop J. (1991) Light limitation of phytoplankton biomass
and macronutrient utilization in the Southern Ocean. Limnol.
Oceanogr. 36(8), 1662–1677.
Monnin E., Indermuhle A., Dallenbach A., Fluckiger J.,
Stauffer B., Stocker T. F., Raynaud D., and Barnola J.-M.
(2001) Atmospheric CO2 concentrations over the last glacial
termination. Science 291, 112–114.
Montoya J. P. and McCarthy J. J. (1995) Isotopic fractionation
during nitrate uptake by marine phytoplankton grown in
continuous culture. J. Plankton Res. 17(3), 439–464.
Moore J. K., Abbott M. R., Richman J. G., and Nelson D. M.
(2000) The Southern Ocean at the last glacial maximum: a
strong sink for atmospheric carbon dioxide. Global Biogeochem. Cycles 14, 455–475.
Moreno A., Nave S., Kuhlmann H., Canals M., Targarona J.,
Freudenthal T., and Abrantes F. (2002) Productivity
response in the North Canary Basin to climate changes
during the last 250,000 yr: a multi-proxy approach. Earth
Planet. Sci. Lett. 196, 147–159.
Morley J. J. and Hays J. D. (1983) Oceanographic conditions
associated with high abundances of the radiolarian Cycladophora davisiana. Earth Planet. Sci. Lett. 66, 63–72.
Mortlock R. A., Charles C. D., Froelich P. N., Zibello M. A.,
Saltzman J., Hyas J. D., and Burckle L. H. (1991) Evidence
for lower productivity in the Antarctic during the last
glaciation. Nature 351, 220–223.
Muller P. J. and Suess E. (1979) Productivity, sedimentation
rate, and sedimentary organic matter in the oceans: I. Organic
carbon preservation. Deep-Sea Res. I 26, 1347–1362.
Murray R. W., Leinen M., and Isern A. R. (1993) Biogenic flux
of Al to sediment in the central equatorial Pacific Ocean:
evidence for increased productivity during glacial periods.
Paleoceanography 8(5), 651–670.
Nakatsuka T., Watanabe K., Handa N., Matsumoto E., and
Wada E. (1995) Glacial to interglacial surface nutrient
variation of the Bering deep basins recorded by delta-C-13
and delta-N-15 of sedimentary organic matter. Paleoceanography 10(6), 1047–1061.
Narita H., Sato M., Tsunogai S., Murayama M., Nakatsuka T.,
Wakatsuchi M., Harada N., and Ujiié Y. (2002) Biogenic
opal indicating less productive northwestern North Pacific
during the glacial ages. Geophys. Res. Lett. 29(15) 10.1029
22.1–22.4.
Opdyke B. N. and Walker J. C. G. (1992) Return of the coral
reef hypothesis: basin to shelf potitioning of CaCO3 and its
effect on atmospheric CO2. Geology 20, 733 –736.
Oppo D. W. and Lehman S. J. (1993) Mid-depth circulation of
the subpolar North Atlantic during the last glacial maximum.
Science 259, 1148–1152.
Pailler D., Bard E., Rostek F., Zheng Y., Mortlock R., and van
Geen A. (2002) Burial of redox-sensitive metals and organic
matter in the equatorial Indian Ocean linked to precession.
Geochim. Cosmochim. Acta 66(5), 849– 865.
Paytan A., Kastner M., Martin E. E., Macdougall J. D., and
Herbert T. (1996) Glacial to interglacial fluctuations in
productivity in the equatorial Pacific as indicated by marine
barite. Science 274(5291), 1355–1357.
Peacock S. (2001) Use of tracers to constrain time-averaged
fluxes in the ocean. PhD, Columbia University.
Pedersen T. F. (1983) Increased productivity in the eastern
equatorial Pacific during the last glacial maximum (19,000
to 14,000 yr B.P.). Geology 11, 16–19.
Pedersen T. F., Nielsen B., and Pickering M. (1991) Timing of
Late Quaternary productivity pulses in the Panama Basin
and implications for atmospheric CO2. Paleoceanography 6,
657–677.
Perks H. M., Charles C. D., and Keeling R. F. (2000)
Precessionally forced productivity variations across the
equatorial Pacific. Paleoceanography 17(3) 10.1029/
2000PA000603.
Petit J. R., Jouzel J., Raynaud D., Barkov N. I., Barnola J.-M.,
Basile I., Bender M., Chappellaz J., Davis M., Delaygue G.,
Delmotte M., Kotlyakov V. M., Legrand M., Lipenkov V. Y.,
Lrius C., Pepin L., Ritz C., Saltzman E., and Stievenard M.
(1999) Climate and atmospheric history of the past 420,000
years from the Vostok ice core, Antarctica. Nature 399,
429–436.
Popp B. N., Laws E. A., Bidigare R. R., Dore J. R., Hanson
K. L., and Wakeham S. G. (1998) Effect of phytoplankton
cell geometry on carbon isotopic fractionation. Geochim.
Cosmochim. Acta 62(1), 69– 77.
Prell W. L., Hutson W. H., Williams D. F., Be A. W. H.,
Geitzenauer K., and Molfino B. (1980) Surface circulation of
the Indian Ocean during the Last Glacial maximum,
approximately 18,000 yr BP. Quat. Res. 14(3), 309–336.
Pride C., Thunell R., Sigman D., Keigwin L., Altabet M., and
Tappa E. (1999) Nitrogen isotopic variations in the Gulf of
California since the last deglaciation: response to global
climate change. Paleoceanography 14(3), 397–409.
Rau G. H., Tkahashi T., and Des Marais D. J. (1989)
Latitudinal variations in plankton d 13C: implications for
CO2 and productivity in past oceans. Nature 341, 516–518.
Rea D., Pisias N., and Newberry T. (1991) Late Pleistocene
paleoclimatology of the central equatorial Pacific: flux
patterns of biogenic sediments. Paleoceanography 6,
227–244.
Redfield A. C. (1942) The processes determining the
concentration of oxygen, phosphate and other organic
derivatives within the depths of the Atlantic ocean. Pap.
Phys. Oceanogr. Meteorol. IX.
Redfield A. C. (1958) The biological control of chemical
factors in the environment. Am. Sci. 46, 205 –221.
Redfield A. C., Ketchum B. H., and Richards F. A. (1963) The
influence of organisms on the composition of seawater. In
The Sea (ed. M. N. Hill). Interscience, Vol. 2, pp. 26 –77.
Reid J. L. (1969) Sea surface temperature, salinity, and density
of the Pacific Ocean in summer and in winter. Deep-Sea
Res. I 16(suppl.), 215–224.
Rickaby R. E. M. and Elderfield H. (1999) Planktonic
foraminiferal Cd/Ca: Paleonutrients or paleotemperature?
Paleoceanography 14(3), 293–303.
Rosenthal Y., Boyle E. A., Labeyrie L., and Oppo D. (1995)
Glacial enrichments of authigenic Cd and U in Subantarctic
sediments—a climatic control on the elements oceanic
budget. Paleoceanography 10(3), 395 –413.
Rosenthal Y., Boyle E. A., and Labeyrie L. (1997) Last glacial
maximum paleochemistry and deepwater circulation in the
Southern Ocean: evidence from foraminiferal Cadmium.
Paleoceanography 12(6), 787–796.
Rosenthal Y., Dahan M., and Shemesh A. (2000) Southern
Ocean contributions to glacial-interglacial changes of
References
atmospheric CO2: evidence from carbon isotope records in
diatoms. Paleoceanography 15(1), 65– 75.
Ruhlemann C., Muller P. J., and Schneider R. R. (1999)
Organic carbon and carbonate as paleoproductivity proxies:
examples from high and low productivity areas of the
tropical Atlantic. In Use of Proxies in Paleoceanography:
Examples from the South Atlantic (eds. G. Fischer and
G. Wefer). Springer, Berlin, pp. 315–344.
Ruhlemann M., Frank M., Hale W., Mangini A., Multiza P.,
and Wefer G. (1996) Late Quaternary productivity changes
in the western equatorial Atlantic: evidence from 230Thnormalized carbonate and organic carbon accumulation
rates. Mar. Geol. 135, 127–152.
Russell A. D., Emerson S., Mix A. C., and Peterson L. C.
(1996) The use of foraminiferal uranium/calcium ratios as an
indicator of changes in seawater uranium content. Paleoceanography 11(6), 649– 663.
Rutsch H.-J., Mangini A., Bonani G., Dittrich-Hannen B.,
Kubik P., Suter M., and Segl M. (1995) 10Be and Ba
concentrations in West African sediments trace productivity
in the past. Earth Planet. Sci. Lett. 133, 129– 143.
Ruttenberg K. C. (1993) Reassessment of the oceanic residence
time of phosphorous. Chem. Geol. 107, 405– 409.
Sachs J. P. and Repeta D. J. (1999) Oligotrophy and nitrogen
fixation during Eastern Mediterranean sapropel events.
Science 286, 2485–2488.
Sachs J. P., Anderson R. F., and Lehman S. J. (2001) Glacial
surface temperatures of the southeast Atlantic Ocean.
Science 293(5537), 2077–2079.
Sanudo-Wilhelmy S., Kustka A., Gobler C., Hutchins D., Yang
M., Lwiza K., Burns J., Capone D., raven J., and Carpenter
E. (2001) Phosphorus limitation of nitrogen fixation by
Trichodesmium in the central Atlantic Ocean. Nature
411(6833), 66 –69.
Sanyal A., Hemming N. G., Broecker W. S., and Hanson G. N.
(1997) changes in pH in the eastern equatorial Pacific across
stage 5–6 boundary based on boron isotopes in foraminifera.
Global Biogeochem. Cycles 11(1), 125–133.
Sarnthein M., Winn K., Duplessy J.-C., and Fontugne M. R.
(1988) Global variations of surface ocean productivity in
low and mid latitudes: influence on CO2 reservoirs of the
deep ocean and atmosphere during the last 21,000 years.
Paleoceanography 3(3), 361–399.
Schneider R. R. (1991) Spätquartäre Produktivitätsänderungen
im östlichen Anfola Becken: Reaktion auf Variationen im
Passat-Monsun Windsystem und der Advektion des Benguela-Küstenstroms. Ber. Fachbereich Geowiss., Univ.
Bremen 21, 1–198.
Schneider R. R., Muller P. J., Ruhland G., Meinecke G.,
Schmidt H., and Wefer G. (1996) Late Quaternary surface
temperatures and productivity in the East-equatorial South
Atlantic: Response to changes in Trade/Monsoon wind
forcing and surface water advection. In The South Atlantic:
Present and Past Circulation (eds. G. Wefer, W. H. Berger,
G. Siedler, and D. J. Webb). Springer, Berlin, pp. 527–551.
Schrag D., Hampt G., and Murray D. (1996) Pore fluid
constraints on the temperature and oxygen isotopic composition of the glacial ocean. Science 272(5270), 1930– 1932.
Schubert C. J. and Stein R. (1996) Deposition of organic carbon
in Arctic sediments: terrigenous supply versus marine
productivity. Org. Geochem. 24, 421–436.
Schwarz B., Mangini A., and Segl M. (1996) Geochemistry of a
piston core from Ontong Java Plateau (western equatorial
Pacific): evidence for sediment redistribution and changes in
paleoproductivity. Geol. Rundsch. 85, 536 –545.
Shackelton N. J., Hall M. A., Line J., and Cang S. (1983)
Carbon isotope data in core V19-30 confirm reduced carbon
dioxide concentration of the ice age atmosphere. Nature 306,
319–322.
Shemesh A., Macko S. A., Charles C. D., and Rau G. H. (1993)
Isotopic evidence for reduced productivity in the glacial
southern ocean. Science 262, 407–410.
527
Sigman D. M. and Boyle E. A. (2000) Glacial/interglacial
variations in atmospheric carbon dioxide. Nature 407(6806),
859 –869.
Sigman D. M. and Boyle E. A. (2001) Palaeoceanography:
Antarctic stratification and glacial CO2. Nature 412, 606.
Sigman D. M., McCorkle D. C., and Martin W. R. (1998) The
calcite lysocline as a constraint on glacial/interglacial lowlatitude production changes. Global Biogeochem. Cycles
12(3), 409 –427.
Sigman D. M., Altabet M. A., Francois R., McCorkle D. C., and
Fischer G. (1999a) The d 15N of nitrate in the Southern
Ocean: consumption of nitrate in surface waters. Global
Biogeochem. Cycles 13(4), 1149–1166.
Sigman D. M., Altabet M. A., Francois R., McCorkle D. C., and
Gaillard J.-F. (1999b) The isotopic compositon of diatombound nitrogen in Southern Ocean sediments. Paleoceanography 14(2), 118–134.
Sigman D. M., Lehman S. J., and Oppo D. W. (2003)
Evaluating mechanisms of nutrient depletion and 13C
enrichment in the intermediate-depth Atlantic during the
last ice age. Paleoceanography 18(3), 1072, doi: 10.1029/
2002PA000818.
Smith H. J., Fischer H., Wahlen M., Mastroianni D., and Deck
B. (1999) Dual modes of the carbon cycle since the Last
Glacial Maximum. Nature 400, 248 –250.
Sowers T. and Bender M. L. (1995) Climate records covering
the last deglaciation. Science 269, 210–214.
Spero H. J. and Lea D. W. (1993) Intraspecific stable isotope
variability in the planktic foraminifera Globigerinoides
sacculifer: Results from laboratory experiments. Mar.
Micropaleontol. 22, 221–234.
Spero H. J. and Williams D. F. (1988) Extracting environmental information from planktonic foraminiferal delta 13-C.
Nature 335, 717–719.
Spero H. J., Bijma J., Lea D. W., and Bemis B. E. (1997) Effect
of seawater carbonate concentration on foraminiferal carbon
and oxygen isotopes. Nature 390, 497–499.
Stephens B. B. and Keeling R. F. (2000) The influence of
Antarctic sea ice on glacial–interglacial CO2 variations.
Nature 404, 171–174.
Stephens M. and Kadko D. (1997) Glacial-Holocene calcium
carbonate dissolution at the central equatorial Pacific
seafloor. Paleoceanography 12, 797–804.
Stutt J. B. W., Prins M. A., Schneider R. R., Weltje G. J.,
Jansen J. H. F., and Postma G. (2002) A 300 kyr record of
aridity and windstrength in southwestern Africa; inferences
from grain-size distributions of sediments on Walvis Ridge,
SE Atlantic. Mar. Geol. 180(1–4), 221 –233.
Suman D. O. and Bacon M. P. (1989) Variations in Holocene
sedimentation in the North-American Basin determined
from Th-230 measurements. Deep-Sea Res. A 36(6),
869 –878.
Summerhayes C. P., Kroon D., Rosell-Mele A., Jordan R. W.,
Schrader H.-J., Hearn R., Villanueva J., Grimalt J. O., and
Eglinton G. (1995) Variability in the Benguela Current
upwelling system over the past 70,000 years. Progr.
Oceanogr. 35, 207–251.
Takahashi T., Feely R. A., Weiss R. F., Wanninkohf R. H.,
Chipman D. W., Sutherland S. C., and Takahashi T. T.
(1997) Global air-sea flux of CO2: an estimate based on
measurements of sea-air pCO(2) difference. Proc. Natl.
Acad. Sci. USA 94(16), 8292–8299.
Takeda S. (1998) Influence of iron availability on nutrient
consumption ratio of diatoms in oceanic waters. Nature 393,
774 –777.
Talley L. D. (1993) Distribution and formation of North Pacific
intermediate water. J. Phys. Oceanogr. 23, 517 –537.
Thomas E., Booth L., Maslin M., and Shackleton N. J. (1995)
Northeastern Atlantic benthic foraminifera during the last
45,000 years—changes in productivity seen from the bottom
up. Paleoceanography 10(3), 545–562.
Thunell R. C., Miao Q., Calvert S. E., and Pederson T. F.
(1992) Glacial– Holocene biogenic sedimentation patters in
528
The Biological Pump in the Past
the South China Sea: productivity variations and surface
water pCO2. Paleoceanography 7, 143 –162.
Toggweiler J. R. (1999) Variations in atmospheric CO2 driven
by ventilation of the ocean’s deepest water. Paleoceanography 14(5), 571–588.
Toggweiler J. R. and Carson S. (1995) What are the upwelling
systems contributing to the ocean’s caron and nutrient
budgets? In Upwelling in the Ocean: Modern Processes and
Ancient Records (eds. C. P. Summerhayes, K.-C. Emeis,
M. V. Angel, R. L. Smith, and B. Zeitzschel). Wiley,
New York.
Toggweiler J. R. and Samuels B. (1995) Effect of Drake
Passage on the global thermohaline circulation. Deep-Sea
Res. I 42(4), 477 –500.
Toggweiler J. R., Carson S., and Bjornsson H. (1999) Response
of the ACC and the Antarctic pycnocline to a meridional
shift in the southern hemisphere westerlies. EOS, Trans.
AGU 80(49), OS286.
Toggweiler J. R., Gnanadesikan A., Carson S., Murnane R.,
and Sarmiento J. L. (2003a) Representation of the carbon
cycle in box models and GCMs: 1. Solubility pump. Global
Biogeochem. Cycles 17(1) 0.1029/2001GB001401.
Toggweiler J. R., Murnane R., Carson S., Gnanadesikan A.,
and Sarmiento J. L. (2003b) Representation of the carbon
cycle in box models and GCMs: 2. Organic pump. Global
Biogeochem. Cycles 17(1) 10.1029/2001GB001841.
Tortell P. D., DiTullio G. R., Sigman D. M., and Morel F. M.
M. (2002) CO2 effects on taxonomic composition and
nutrient utilization in an equatorial Pacific phytoplankton
assemblage. Mar. Ecol. Progr. Ser. 236, 37–43.
Tyrrell T. (1999) The relative influences of nitrogen and
phosphorus on oceanic primary production. Nature 400,
525 –531.
Vincent E. and Berger W. H. (1985) Carbon dioxide and polar
cooling in the Miocene: the Monterey hypothesis. In The
Carbon Cycle and Atmospheric CO2: Natural Variations
Archean to Present (eds. E. T. Sundquist and W. S.
Broecker). American Geophysical Union, Washington DC,
pp. 455–468.
Volk T. and Hoffert M. I. (1985) Ocean carbon pumps: analysis
of relative strengths and efficiencies in ocean-driven
atmospheric CO2 changes. In The Carbon Cycle and
Atmospheric CO2: Natural Variations Archean to Present
(eds. E. T. Sundquist and W. S. Broecker). American
Geophysical Union, Washington DC, pp. 99–110.
Walsh J. J., Rowe G. T., Iverson R. L., and McRoy C. P. (1981)
Biological export of shelf carbon is a neglected sink of the
global CO2 cycle. Nature 291, 196–201.
Walter H.-J., Rutgers van der Loeff M. M., and Francois R.
(1999) Reliability of the 231Pa/230Th activity ratio as a tracer
for bioproductivity of the ocean. In Use of Proxies in
Paleoceanography: Examples from the South Atlantic (eds.
G. Fischer and G. Wefer). Springer, Berlin, pp. 393 –408.
Warren B. (1983) Why is no deep water formed in the North
Pacific? J. Mar. Res. 41, 327 –347.
Waser N. A. D., Turpin D. H., Harrison P. J., Nielsen B., and
Calvert S. E. (1998) Nitrogen isotope fractionation during
the uptake and assimilation of nitrate, nitrite, and urea by a
marine diatom. Limnol. Oceanogr. 43(2), 215 –224.
Watson A. J., Bakker D. C. E., Ridgewell A. J., Boyd P. W.,
and Law C. S. (2000) Effect of iron supply on Southern
Ocean CO2 uptake and implications for atmospheric CO2.
Nature 407(6805), 730 –733.
Weaver A. J., Eby M., Fanning A. F., and Wiebe E. C. (1998)
Simulated influence of carbon dioxide, orbital forcing and
ice sheets on the climate of the Last Glacial Maximum.
Nature 394, 847 –853.
Webb R. S., Lehman S. J., Rind D. H., Healy R. J., and Sigman
D. M. (1997) Influence of ocean heat transport on the climate
of the Last Glacial Maximum. Nature 385(6618), 695–699.
Wefer G., Berger W. H., Siedler G., and Webb D. J. (1996) The
South Atlantic: present and past circulation. Springer,
Berlin.
Wollenburg J., Kuhnt W., and Mackensen A. (2001) Changes in
Arctic Ocean paleproductivity and hydrography during the
last 145 kyr: the benthic foraminiferal record. Paleoceanography 16, 65–77.
Yu E. F., Francois R., Bacon M. P., Honjo S., Fleer A. P.,
Manganini S. J., van der Loeff M. M. R., and Ittekot V.
(2001) Trapping efficiency of bottom-tethered sediment
traps estimated from the intercepted fluxes of Th-230 and
Pa-231. Deep Sea Res. I 48(3), 865– 889.
q 2003, Elsevier Ltd. All rights reserved
No part of this publication may be reproduced, stored in a retrieval system or
transmitted in any form or by any means, electronic, mechanical, photocopying,
recording or otherwise, without prior written permission of the Publisher.
Treatise on Geochemistry
ISBN (set): 0-08-043751-6
Volume 6; (ISBN: 0-08-044341-9); pp. 491– 528